The roadster predates the Leaf but it was very expensive. The Leaf was the first affordable EV and in that they beat Tesla by a couple years.
Anyone laughing at what SpaceX was doing was ignorant of things like the DC-X. I suspect a fair amount of that laughter was actually fear from a moribund industry happy with the status quo rather than actual skepticism that it could be done.
Some skepticism about a company the size and budget of SpaceX being able to do it might have been reasonable. That is probably the most impressive thing. The DC-X was Lockheed.
I think SpaceX is meaningfully different than Tesla; even now, Musk seems to display a substantially different personality with regard to SpaceX vs. everything else — the similarity is the optimism, sure, but in he's a lot more willing to say "this is hard and I will make mistakes".
As for how hard the SpaceX stuff is: all I can do there is look at all the rocket companies and space agencies that were openly laughing at what SpaceX was proposing to do. I think SpaceX is where the market proof is for that, rather than the other way around.
Based on what I've heard from civil engineers and neuroscientists, with regard to TBC and Neuralink they regard him much as the annoying speaker with no self-awareness in https://xkcd.com/793/
FSD sounds like what happens with time estimation for someone new to software engineering projects — you have to take what they say, double the number, and increase to the next highest unit: "1 day" means "2 weeks"; "3 months" means "6 years"; and if someone says "next year" in 2016, pencil 2036 into your diary.