How good is $20K seed capital?

I've started doing rounds to raise seed capital from friends/family - for a SaaS product idea. The idea expects investment of $20K and extrapolates that to multi-million dollar company in 3-5 years.

Few people have "questioned" if that is a good seed amount, and below is the conversation from one such case.

Does anybody see anything wrong with my approach or assumptions? (as detailed below)

Please bear with me if I am writing something totally stupid - I've never raised seed capital earlier.

The conversation between ME and JOE goes like this. JOE is not an angel investor nor a VC. JOE is a friend who has the seed amount I require

---------------------
---------------------

ME: I've a product idea that can generate a revenue of $500K per year after 1.5 years. And that it can reach $1m within 3 years. And can scale to a billions in 5-7 years if we strategically take it  further.

(Further, I say)
We create a company and you(read: the other party JOE) get 40% of the equity. But you must pay the complete seed amount to the company initially. And the company pays me a salary XYZ from the seed amount for 1.5 years. i.e it goes to company's expenses. (The salary amount is needed to meet my monthly expenses)

JOE: Okay, I hear you. What is the amount you are looking for?

ME: $30K is good, $40K is excellent, $20K is okay - but not below $20K.

ME: You should also understand that if the idea fails, you lose the amount equal to the amount in company's expenses.

JOE: Really - 20K? Are you sure about your costs and your projections?

ME: I'm sure about the costs. I have never started a company before but the estimates after 3 years are pretty conservative from my perspective.

ME: Let me explain the idea and the plan
...I explain him the idea and he is pretty convinced that it is a workable one. However, he still looks at the business model as "too good to be true"

-----------------------

Some of the points that I mentioned to JOE about the plan

1. I live and work from India. My expenses are lesser than western counterparts (lets say 1/2). i.e a small capital goes a long way.

2. I'll be working solo for long time in first 1.5 years- because I can do work of multiple average-to-good programmers/QA alone (no exaggeration). NOTE: Good != Great

3. There are certain things I cannot do myself - but I know freelancers who will do it for a fee.

4. I am expecting the product to market itself, and therefore does not requires heavy sales/marketing - atleast not in the early stages (1.5 yrs)

5. The development plan is something like this:
 --- Core framework (value enabler)- to be finished in max 3 months
 --- Value creation stuff, beta releases - between 4-6 months
 --- GA after total of 9 months (with higher focus on customer acquisition and we probably approach VCs for further funding after this milestone)