Hello Richard,
Thanks for the reply. There’s more there than I can quickly respond to, and I’ll be busy for the next few days. We might end up talking past each other for a bit, as I am still working on the second part of my reply.
Some of the citations you’re looking for are just working familiarity with published data sets. Perhaps @glipsnort could also weigh in - he has discussed the allele frequency distribution here previously (and with nice graphs). The human allele frequency distribution as a whole is one very good piece of evidence that we do not come from just two people in the last few hundred thousand years. You could try messing about with a starting pair and mutation frequencies and see for yourself the challenge of generating the distribution we observe.
Also, keep in mind you’re asking about a bottleneck to two - not 2,000, not 200, not 20 … you get the picture. Moreover, you’re asking for a census size of two, not just Ne =2 . It’s a long way down from thousands to 2.
More anon - thanks as always for your patience.