The argument: simulating weather is super hard because weather has a "butterfly/snowball effect." Small perturbations in initial conditions result in large difference in later states of the simulation. Missing some precision in modeling the initial conditions or rules of the system is equivalent to having lots of little butterfly wings flapping all over your simulation. The brain is even more complex than a weather system. Therefore simulating a brain will be even harder than simulating the weather.

There are at least two problems with the argument:

(1) The goal of simulating the weather is to get the simulation to match what will happen in reality as closely as possible. The goal of simulating a brain is to get a virtual brain that is about as smart as a human. If the simulated brain is as different from the human brain it was simulated from as two human brains are from each other, it doesn't matter. If a weather simulation is as different from what actually happens as two similar weather systems starting from similar initial conditions can be, it's a failure.

(2) We don't know if human brains are fragile enough to have "butterfly effects" that matter. Just because a brain is complex doesn't mean that it must have these butterfly/snowball effects. Human brains are famous for compensating for damage, and still functioning pretty well when damaged.