But that's not a sufficient counter-argument that the GGP post's reasoning is sound. Even single base pair mutations can have huge impacts. A surprising percentage of the human genome is retroviral in origin. As far as I know, "Genetic clock" analysis is the only sound application of counting base pair differences.

I don't think the GP is arguing one way or the other. They seem to be mostly questioning how wide the error bars should be on these statements. It looks like a Socratic twist on "citation needed" without an implication that statements are necessarily wrong.