> All non-African populations diverged from each-other more recently than their shared common ancestor did from their ancestral African population
That's a much weaker claim than earlier, more specific out-of-Africa theories that claim there was a very simple, very linear migration without any possibility of genetic back flow. And that's ignoring the fact that it ignores archaic humans. Despite being backed up with supposedly irrefutable genetic models, in recent years the dominant out-of-Africa hypothesis has become weaker as it's had to contend with, e.g., proof of Neanderthals and more recently Denosovans interbreeding that directly contradicted the predictions of previous models.
I'm not any kind of trained scientist, but I have studied enough to know that while the dominant models (the specific ones--Jordania is not contending that humans didn't emerge from Africa) absolutely pose a signifiant hurdle for Jordania's hypothesis, they don't rule it out and, in fact, since he began constructing the model some 30 years ago (formal publication of the theory with a comprehensive presentation was about 10 years ago, I think) the dominant models have become increasingly more accommodating.
Regarding Asian music, Jordania is an ethnomusicologist. The bulk of his paper, book, and other papers discuss his model from the perspective of ethnomusicology. It's strongest in its discussion of, e.g., phonology. It's weakest from the perspective of accepted anthropologic genetic models of evolutionary development. But as I said what makes it most compelling to me is how it solves (at least in the abstract) the free-rider problem squarely within the confines of traditional, well-proven genetic theory, through an intuitively simple and obvious (in retrospect) survival adaptation that knits together several inexplicable human behaviors. That's hugely significant given, AFAIK, no other theory of human evolutionary development can make that claim. There'll all predicated on some unknown or unproven extension to standard genetic models (e.g. group selection). Of course, such elegance alone isn't proof of anything. But he does provide a cogent argument with substantial evidence, albeit predominantly deriving from the field of ethnomusicology. And as I said, genetic anthropology has slowly become more favorable, or at least less antagonistic.
I think scientists draw too many and overly specific conclusions about evolutionary development from models of genetic drift, etc, which is why we see the backtracking as the field adjusts to facts that would never be predicted by or outright contradict those conclusions.[1] In any event, even in Jordania's model all the physiological traits and most of the psychological pieces could have been (and presumably were) already present even at the first migration out of Africa. His argument about back-migration of the allele for articulated speech isn't essential to the model in terms of the order of events. It's just the most easily tested aspect and is emphasized partly because of how he arrived at the model--as an ethnomusicologist following the trail of evidence he was most familiar with. I can easily see the disparity in, e.g., stuttering disappearing with better data; certainly he admits the paucity of comprehensive studies.
It's worth at least skimming his original thesis paper or the book, available on Amazon Kindle. It's admittedly radical and not well known, but IMO substantive and rather fascinating. I've read the whole paper (in piecemeal fashion) but am not an able defender given my lack of expertise.