To be clear, I only noted Genghis Khan (who alone has some 32 million+ descendents) to demonstrate how a "vanishingly rare" trait can become extraordinarily common if the right person happens to have it.
Let's hypothesize two "vanishingly rare" traits floating uniformly around the primordial African populace. As humans migrate out, some equally "vanishing" proportion carry said traits out along with them. Then, for whatever reason, the "vanishingly rare" traits begin propagating at a higher rate. Once the humans have migrated over whatever range, that pivot can take place anywhere: i.e., all else being equal, it's just as likely to take place in Fujian as it is to take place in Axum.
So, if one discounts pure chance, the question you're left with is why it would appear in Asia and not in Africa. I can throw out one hypothesis right off the bat: The Toba eruption 73,000 years ago induced both a crash in the global population and severed the Southeast Asian remnant from the ancestral African homeland. This may be nothing more than a chance disproportionate survival of these two traits in that now-isolated population as compared to the surviving humans that had remained in Africa.
And that's just one random conjecture. The ultimate point that I'm advancing is that you should see a much more dramatic divergence if non-Africans were a hybrid of Homo sapiens & Homo erectus whereas Africans remained 'purely' Homo sapiens. That's what this hypothesis is suggesting and I just don't think there's anywhere near enough evidence to support that conclusion.
There was an explosion of human 'branches' at this time and then again during/at the Last Global Maximum, 25-18,000 years ago. (Oppenheimer)