I was skimming through Glenn Conroy's Reconstructing Human Origins today, and it was like I was reading John Hawks' weblog. Conroy spends a good chapter ripping into simplistic conflations of gene genealogies with the evolutionary history of a species. He hits points like the likely non-neutrality of mtDNA & Y lineages, the affect that different population sizes have on phylogeographic reconstructions and the loci which fly in the face of a simple out of Africa model. Here he is in plain English:

...when dealing with populations that can interbreed, morphological replacement should not be equated with population replacement.

Like I say, morphology does not imply phylogeny. Unfortunately, the mixed "mostly Africa" model that Conroy seems to lean toward escapes pithy sloganeering and summation in the manner that Out-of-Africa and Multiregionalism seem amenable too. Out-of-Africa's one shot expansion from a fixed space and time to fill up the world is pretty easy to conceive of. Similarly, Multiregionalism as old fashioned anagenesis could be conceived of as a flux of genes amongst a node of fixed populations. The reality might very well be a palimpsest of nodes overlain with important demographic expansions from particular locales, along with the important caveat that regionally adaptive genes might flourish indefinitely while ancestrally informative alleles sink or swim in the soup of migration, mutation and drift. Oh, and while I'm at it, "Nature vs. Nurture" is an idiotic dichotomy.