Okay, maybe there aren't other examples quite as good as him, but a few of these people surely come close.

Or we could simply point out that with average IQs in the 70s and 80s, average mathematician IQs closer to 140s - or 4 standard deviations away, even in a population of billions we still would only expect a small handful of Ramanujans - consistent with the evidence.

Yes, but I'm not sure all of the populations working in cotton fields and sweatshops had such a low average IQ. (And Gould just said “people”, not “innumerable people” or something like that.)