My problem with Mark's thinking here is that we're all mutts. You might have certain genes that are markers for a certain geographic location, but who says that necessarily translates to food tolerance/intolerance? I might buy it if you can trace your lineage as a "purebred" for many generations, but that probably doesn't happen for most humans. We immigrate, we flee, and our armies conquer each other and make babies with other populations.

Analogy: I'd find it plausible that Dobermans have certain nutrition needs and problems, and Huskies have different ones. Purebred dogs are a genetic bottleneck. But a mixed breed that kinda seems to be part Doberman, or Husky, or any other breed? How do you know which nutrition genes they inherited, unless they're specifically linked to the other marker genes that are being identified?

Essentially, I'm skeptical for the same reason I'm skeptical on the blood type diet. You'd have to demonstrate that those ancestral gene loci and whatever nutrition gene loci are close together, meaning there's little crossing over on chromosomes.