One peer reviewed article would probably do the job, decent ones do not let unreferenced claims through. I hope you saw my edit, I did read the paper a bit.
I need to actually get work done at some point today, so I'll see what I can do later on. Also, if you find some more studies about clustering coming, I'll take a look at them.
No one is actually disputing that claim. What we are disputing is concluding from those genetic differences that therefore English and Native Americans are members of different "races."
When people like Reich hear "race is just a social construct" it seems they hear "meaningful genetic differences don't exist between English and Native Americans." And then they respond saying "no, race isn't just a social construct." Then you say "but if we look at the genealogy of the concept of race, you'll find its origins rest on historical contingencies surrounding power dynamics, colonialism, and the 17th century political economy. Plus the classic racial categories are just kinda bs." And then they think "well, yeah, but that's not really what I'm talking about. I'm saying meaningful differences are correlated with where a person's ancestors were predominantly located 500 years ago because before that all of their ancestors had been separated for maybe 40,000 years. So the English and Native Americans are in fact divergent populations..." And then someone's like "so are you saying those are two different races??"
I'm not clear on where your goalposts are or if we're even in disagreement about anything besides what we think David Reich is trying to say.
Then why is he saying things like
But as a geneticist I also know that it is simply no longer possible to ignore average genetic differences among “races.”
A chapter of his book is devoted to the vast genetic diversity within Africa. He is the source for which I claim huge amounts of genetic diversity exist within Africa. E.g., huge differences between the San, Pygmies, and Bantu. He's well aware that defining a "black" race is problematic and this quote doesn't show he tries to do so. He puts the word "race" in scare quotes for a reason.
What he is saying is that (1) genetic differences exist between populations whose ancestors were separated by 70,000 years prior to colonization, (2) at some level of granularity, e.g., African Americans are a population of relatively similar ancestry (note the US does not have a lot of African pygmy or San peoples) and this ancestry is separated from, e.g., European ancestry by like 70,000 years, and (3) tens of thousands of years is long enough for SNPs to build up. And for that reason, we observe differences in the distributions of medically relevant phenotypes, like propensity for a number of diseases between self-identified African Americans and self-identified whites. Which parts of this, exactly, do you object to?
My guess is you object to (2). But I don't think his goal is to say there is some globally optimal level of granularity or height of the hierarchical clustering tree we should always stick with. Some mutations, like adaptations against malaria, are probably pretty high up the tree, in which case it makes sense to discuss the subtree of that height. I highly doubt his hill to die on is arguing we shouldn't adjust the granularity in different situations. His hill is that the tree matters and his argument is that people refuse to believe any such hierarchical tree can or should ever be created. I imagine he's also kind of taking for granted that the granularity already exists in people's identifications and given that granularity, we do in fact see meaningful differences between those subtrees and other parts of the overall tree. If African Americans had the luxury of identifying with particular parts of Africa, maybe it'd be different.