I think my last comment may have implied a bit of disdain for the quality of talent in non-SV places. That isn't the case. I agree with you, I think, that the number and quality of talent outside of SV is sufficient to sustain almost any startup in theory. In practice, purely for artificial reasons, it doesn't hold.

There are two barriers to reaching the "critical mass" you note outside of the valley:

1. The "critical mass" in SV of "interesting," "good" (or whatever favorite positive adjective might apply) is just huge. Even the least interesting corporate-y jobs are strongly influenced by the culture here, and that is a big draw for talent from bottom to top. That's all I really meant, earlier: the density of "good" jobs here is very large.

2. SV is (over-?)run with people who--and this is another of my unpopular opinions around here--have huge egos. I mean spectacularly huge egos. I come from a background of academic research (very mathematical in nature). Yet I never saw such egos as I've seen in this industry. The barrier here is something of a self-imposed psychological one, I think: the fact is that the vast majority of startups don't need even a single person who has an advanced, in-depth, and broad understanding of mathematics, data structures, algorithms, or basically anything else in academic CS, but they seem to treat it as the minimum necessary element to hire. The armchair psychologist in me opines that the interview process in this industry is more about certain people trying to impress themselves or prop their own egos up instead of just accepting that working on some shopping cart, ad analytics platform, or generic-social-media-app-X just isn't that complicated (speaking academically/intellectually here). They don't want to think they're working on something that is academically trivial, and so they engage in employee selection processes that reinforce the view that everything they do is super intellectual and requires only the best minds from the CS world to succeed.