First off, I have tons of game design experience and with marketing. Everything presented to me in recent times seem to show that OW is shifting to a F2P model.

Games don't have to be good to sell. It's all about marketing and your model. A game like Titanfall 2 which is deemed perfect for a new IP failed because it's bad marketing and artstyle. If I was in their position I would make it an anime FPS game. This is often what the Koreans/Chinese do to failing FPS games with them just replacing all artwork with anime girls. Strange isn't it?
Lest not forget that the biggest Patreon content makers with 20,000$ average income all sell anime/hentai/patreon games whilst serious projects ones barely reach 800$ average.
Examples: Yandere Sim, Monster Island RPG, some porn games and Visual Novels.


So to answer the question, will Overwatch head towards F2P?
Most likely, yes.
Which will be explained below:

Overwatch has a serious problem. It's dwindling playerbase and fatigue.
Currently everything that is happening in OW on the development side seems to be pushing for F2P.

I will note a few cases which have popped up as odd.

1. OW does not have an e-sports scene to carry it PR side.
Believe it or not, e-Sports in OW is quite unpopular. The biggest complaint is that it's impossible for spectators to follow and also understand what's going on.

Due to the mass amount of particle effects, abilities and of course damage outside of camera range it's often hard to tell what is going on. An experienced player will obviously understand but not to the average Joe sitting on 480p stream with the e-sport caster shouting down the mic.

And of course nauseating motions that happen are also one of the big hinders.

Games who are/were carried by theIR e-sports side:
SCBW
SC2
CSGO
DOTA 2

2. Losing streaming space.
For a "growing" game, it's unfortunately losing viewership on Twitch.
Games like PUBG are overtaking its space due to being more eye friendly and intense.

Example
Streamer content:
Overwatch has been streamed for 1 year, was a huge favorite for the first two months.
Total streaming time: 20 years, 362 days
Playerunknown's Battlegrounds, released almost 2 months ago.
Total streaming time: 27 years, 297 days

Audiences' time spent watching said streams:
Overwatch (1 year):
507 years, 85 days
PlayerUnknown's Battlegrounds (2 months):
1,224 years, 214 days


Note the disrepancy here?
OW has almost the same amount of content output, but no interest from the audience.
This means if there were 3 people deciding between a stream, 2 would go for PUBG.

Okay, I am just joking. The real scenario would be:
Overwatch (1 month):
42 years.
PlayerUnknown's Battlegrounds (1 month):
612 years.

Out of 15 people or so, only 1 would tune into an Overwatch stream.
The rest would go to the alternative FPS PUBG.

3. Mobile game design aka "Burst" gaming.
This was coined by Hideo Kojima I believe. When he designed MGS Peace Walker he took into account that a lot of people that play on the handhelds like PSP only spent 10-15 mins or so. This means you could finish a mission on the train or in a waiting room.

A game that is played in short bursts will lose interest real fast, this is more of a casualization of a game. Requiring less time and effort to play and of course concentration.
Example of such games: Candy Crush Saga, TempleRun and a huge lot of Nintendo DS games (Rhythm, Puzzle etc). All which have a timespan between 1-5 minutes.

Non examples of such cases: DOTA 2 (40-50 MIN AVG), CSGO (30-90 MIN AVG), PUBG (25-30 MIN AVG)

Now let's run through Blizzard games:
HOTS: 10 MIN - F2P
HS - 7 MIN - F2P
OW - 10 MIN - P2P

Less time spent means less potential in-game sales!
But obviously HOTS/HS bring in money, why shouldn't OW?

If your average Joe, spends 10 mins in a match and then quits the game that's less chance of a conversion (monetary transaction) compared to Michael who will sit in a CSGO/DOTA2 game and look at someone else's cool skins for 40 mins before craving them himself.

The other problem is fatigue. If a game length is 30 mins, it means the outcome is already decided and it will follow a very linear path of what will happen in said game. This means the player has to play 3 more matches before he gets enough and has seen it all, for the day.

However in OW you will join 3 games within 30 mins and have experienced stomps/griefing/carrying/different players/playstyles at much faster rate. This exhausts the gamer or rather appeases him and he gets enough for the day.

The only solution is to scale it with huge amounts of player growth and then just bank on them purchasing lootboxes. If you get a report that 1 out of 10 players are buying lootboxes, you cant rely on the slow growth of the game. You'll need a quick way to have 2 out of 20 players buying lootboxes. Thus a F2P solution is optimal.

4. Casualization in general.
Here's important parts you'll need to know, this is mostly game design and can be applied to a lot of games.
But everything listed here is to drag any F2P player and have them keep playing your game even though they don't understand anything.

A. Game features a helper to help a casual player perform better.
Examples are in-game macro functions, clearer visuals, aim-assist, more team oriented to have less load on 1 player.

Examples of this:
Quake Champions (RED OUTLINES, AUTO BHOP, SHOOT-ME SIGNS, TANK CHARACTERS FOR LESS SKILLED PLAYERS) - F2P
Has lootboxes: Yes

PALADINS (RED OUTLINES, TUTORIAL, HELP LINES, TANK CHARACTERS, NON-AIM TYPES CHAMPIONS) - F2P
Has lootboxes: Yes

OVERWATCH(RED OUTLINES, TUTORIAL, HELP LINES, TANK CHARACTERS, NON-AIM TYPES CHAMPIONS)
Has lootboes: Yes
Non F2P

GLOBAL OFFENSIVE ()
Has lootboxes: Yes
Non F2P

TF2 (TUTORIAL, "CHAMS" TYPE CLOTHING, VISUAL DMG GIVEN, TANK CHARACTERS) - F2P
Has lootboxes: Yes

As you can see all F2P games share the same thing. Over casualisation.
But CSGO has Lootboxes, why not go F2P?

Well first, game is super cheap. Second is the lootboxes (and their content) can be sold and resold which gives Valve an extra cut meaning the playerbase is essentially working for Valve.
Third reason is the higher skill gap meaning having a F2P base wouldn't do the game much help as most players would quit due to how hard the game is and unforgiving. It would just be a drain of resources.

B. Ranked is not being focused on.
If the playerbase is not encouraged to compete then it's a casual game. Jeff Kaplan has confirmed it recently that he detests giving rewards to ranked players and rather want people just to have fun and enjoy the game rather than just playing competitive.
This is quite the statement that contradicts what the playerbase wants:
Example:
OW Ranked was meant to have longer off-season periods. 1,5 months to be precise.
However due to the huge population drop after ranked ended, they brought it back immediately.

Ranked ELO means nothing and is just to make you feel good and is quite forgiving:
Example:
In most games you have 1 rank. In casual games you often have multiple ranks so that you can "feel" good about yourself. This is common in F2P games, especially the P2W mobile games.

No match history is allowed. This was explained that they did not want people to look at their past performance and get angry. This is more of a brainwashing and to make the player feel good about themselves rather than sulking over 10 losses in a row.

5. Testing the waters Part 1

First off, I have to bring this up before explaining some interesting stuff later.
OW has 30 million copies sold. However did you know that that only nets them 20% of the profit OW gives to Blizzard?

That's right. 80% comes from Lootboxes.
Blizzard has recently briefed their investors that they get 80% of their 1 billion dollar revenue from lootboxes which means that investors will force them to focus on solely lootboxes.
This happened to Call of Duty franchise which Blizzard and Activision own, Bobby Kotick has explained that's solely the reason why Call of Duty is the way it is today and that they pretty much mapped future games from these sales calls and brainstorming how to scale.

You can bet your ass they just did that with Overwatch a week ago.

5. Testing the waters Part 2

This is already happening before your eyes, OW seems to be headed towards F2P market.

Examples:
Kaplan stated there would be 4 season events at most and wanted to focus on new maps and heroes.
Today there are 6 events with the anniversary ones. There are further ones planned.

It's been recently revealed that most of their development team is working on events themselves.
This is worrisome as an old development interview revealed they needed 3-4 months to release 1 map and would often work on 3 others and with 2 often being scrapped. If you transfer workload to these event ones that means less maps and more recycling.

And not surprisingly 3 arena maps have been revealed, that have recycled materials and are in same areas as previous maps. This is contradictory to their "diverse" landscape they were promising with us being taken all over the world with each new map.

Endless rotation of events.
Currently their plan is to have enough events in OW so they can abandon it and just have the game running every year with these events on rotation. This means every 2-3 months they would get a huge sales boost from an event due to F2P crowd being enticed into investing into these exlusive skins.

Rotating events are worthless to old players.


Revival of an event.
The most alarming one was the Uprising event being brought back from the dead after it had run out, but you guessed it, they didn't bring back the event itself. But the Lootbox purchase option for 48 more hours.
This is either a sign that they were underperforming in sales or rater just wanted to fetch more money which no doubt would have/be suggested in the performance call of Blizzard/Activision.

Free-2-Try events will grow more.
Usually Free-2-Try events are given to new but more frequently older games that have a dwindling population. This is obviously to encourage players to buy their game or foreshadowing a F2P conversion.
During these Free-2-Try events, the devs will be tasked with monitoring population numbers. If it is shown that the games population grows 300%, then the game turns into a F2P model pretty fast.


No sequel option.
This is the biggest problem. Currently games like CoD or any other yearly game such as Battlefield series cannot go F2P even though both have extensive lootbox systems, mainly because they get a sequel released almost yearly meaning if the population dwindles they only need to release a new game and they're set.

OW on the other hand cannot do that, it's a base game meant to be endlessly upgraded with content. It would make no sense to release OW2 when the game itself is polished and just needs more content to entice players to remain.
Thus along the road they will be forced to turn to F2P.

6. Wardrobe.
Game is turning into TF2. Emotes, dancing, more jokey cosmetics such as Bastion racercar, tons of interactivity and customization is what I'd expect from TF2.
Not only that but the lootbox system is designed to drain people of their money with the duplicate and sprays filling up the roulette wheel.
This is all designed for F2P players and not people who spent 50$ already for the base game. It's just too expensive but not for F2P who will try their luck and then stop after 15-20$ but in greater numbers.


7. OW has the fastest 'drop off' curve.
This is often to show how fast a player learns what makes the game bad and quits.

In CSGO it takes around 300 hours to get a hang of the game and understand game mechanics, OP weapons or strats. Just a general idea. This is also where people can choose to quit. With 700 hours being average game time before a player quits the game indefinitely.

DotA 2 requires 800 hours to get a hang of it and 2000 hours being the average before a player will quit indefinitely. In some areas it's much much higher due to PC gaming culture.

OW has around, I'd say 14-20 hours. This is based on all of my accounts (I've had 5) and the max amount of hours I saw on my friends list being 150 hours. On launch, I had 20 friends. After 2 weeks, only 3 logged in to play. I'd argue the super fast games and the horrid game design is behind it.
Example being: Roadhog hook frustrating players, this means a lot of people will just stop playing because they know Roadhog is a problem for them and they don't want to continue playing if he's gonna be in every game.

It takes 1-2 games to understand Tank meta is a problem, 20 mins and you have a lot of expertise on what makes the game's meta bad.

Non-rewarding heroes. This is another design flaw. Having a hero like Bastion whose turret can just spray relentlessly at enemies with no game skill required except moving your mouse slowly around the target hero.

Soldier has no recoil, meaning a player has no need to control it. Same goes for the Bastion-esque gameplay by just pressing Q and winning.

The other high reward heroes get nerfed meaning skill gap decreases. A common argument is that nerfs = harder to play = higher skill gap.
WRONG. Timing is everything. It's understandable that you can perform bad with a shit hero but on the other hand if you rely on performing well by learning timings, rotations, skills, counters and then the developers nerf them or just straight out remove them then you lose everything.
Imagine if Lucio's wallriding was time based for 3 seconds. Wow, now professional players cannot perfectly time between wall jumps, execute strats, do advanced bhopping requiring walls and so forth.

However I will tell you that if you see a developer nerfing a good mechanic, then they are guaranteed casuals that don't play the game and rely on pro player gameplay and just think for themselves "he couldn't have gotten 1000 billion kills if we removed that thing he did at the 0:34 mark!"
Passionate players often compromise with more reasonable nerfs but not mechanic based ones. One example being the Genji walljump or double jump reset removal.

Final notes.

When OW's success was announced, with a 1 billion dollar revenue, it meant that the game would be required to keep up the performance. This is why things like Overwatch League are being pitched. They want the game to remain popular.
Kaplan himself said he doesn't want to split the playerbase with premium maps or premium heroes (or rather pay to unlock) which suggests having a F2P model with everything unlocked except lootboxes makes sense.
Which is their primary goal. To get people to buy lootboxes.