I'm waiting to see some good statistical analysis applied to a sample of diverse human genomes.
I have never personally operated a PCR machine so what I say must be discounted appropriately.
I know from reading that there are roughly 20,000 genes that are associated with specific protein generation in the core DNA.
What I know has not been done, so far, and which would make a great deal of sense to me would be to take a sample of different human populations determine the sequences that code for their roughly 20,000 protein encoding genes and do an analysis of variance. A matrix with different individuals on one dimension and their sequenced alleles in the other dimension.
An analysis of variance would show us what are the deeply abstract differences among human populations. We know there areanatomical differences that have been carefully recorded ranging from soft versus hard earwax, large butts versus small butts, melanin differences, lactose processing differences, sweat and follicle differences, etc.
What we don't know is how significant these clusters of differences are. Does it take many genes to create these differences or only a few. That in itself would be a clue as to the number of years of evolution involved in human differences.
An analysis of variance would show first the most significant differences. The largest number of protein encoding genes would undoubtedly have to do with sexuality and reproduction. Those are often the major categories of DNA in other species especially microscopic species. Sexual differentiation defines the species.
Beyond this initial cut, it would be interesting to see whether our visual perceptions tell us about real genomic differences. We tend to sort people on the basis of their melanin level, their hairstyle, their epicanthic fold and their noses. I seriously doubt that the initial analysis of variance would find these to be major population differences.
I hope someone will have the data available and the ability to do the statistics.