Most intergenic mutations have less effect even than synonymous mutations. In any case, it doesn’t matter: mutations with very slight effect on fitness will not be selected against, but they also were never selected for, so most sites in that class will be more or less at equilibrium, with as many very slightly deleterious mutations as very slightly beneficial, all of them meaning nothing in the long run.
Nah, that’s quite incorrect. Natural selection can remove an arbitrarily large number of deleterious mutations every generation. The limiting factor is how large a burden of deleterious mutations an individual can carry. Suppose the average number of deleterious mutations carried by people is 1000. That’s just the average, though. Some will be carrying 800 and some 1200, for example. The ones with 1200 will be a lot less likely to reproduce than those with 800. For every 800er that reproduces in the place of a 1200er, effectively 400 deleterious mutations have been removed.
The only assessment they make in that paper of neutral mutations is for amino-acid altering ones, which is a tiny fraction of the total. The paper has no relevance to the discussion at hand.
A 6000 year long history of humanity, starting with a single couple, is completely inconsistent with observed genetic variation in current humans. You cannot get the genetic variation we see, in the patterns we see it, in less than about half a million years. We’ve been over this at excruciating length here previously, in this thread