Dienekes Pontikos has a long post up on how reticulation within phylogenetic trees may distort our perception of human natural history when we force the data into a more conventional tree (i.e., bifurcation after bifurcation). The concrete reason for this rethinking is the high probability of "archaic admixture" into the dominant genetic signal of anatomically modern African humanity, both within Africa and outside of it. Dienekes proposes that when ancient DNA from early modern Eurasians is analyzed then a large portion of the portrait will be unmasked. For example, if high levels of admixture were present very early on then you would see very divergent regional populations because of persistence and continuity of local hominin population substructure. The pre-African Eurasians from each given region would have contributed substantially to the genetic makeup of the first modern humans who flourished in Europe and East Asia. On the other hand, if admixture was minimal, then the early Europeans and Asians would be far less distinct than their modern descents. This is where I want to highlight one aspect of Dienekes' model which is implicit, but I think needs to be strongly emphasized. The ancestors of modern Europeans and East Asians may not, and in fact I do not believe they are, predominantly descendants of the first settlers of Europe or East Asia. There are already hints of this as late as the Bronze Age, and I think it must have been substantial during the Ice Age as well. Perhaps less due to demographic replacement as meta-population dynamics, as local demes may have regularly gone extinct, so the "periphery" may have been settled more the denser "core" repeatedly.
A few years ago the model was that modern humanity replaced archaic lineages ~50,000 years ago. Part of that replacement was due to the emergence of 'behavioral modernity,' as distinct from 'anatomical modernity,' which predated the former by tens of thousands of years. That behavioral modernity was most clear in evidence in western Europe, in the artistic explosion of the early Upper Paleolithic. There is now a high probability that genetic contribution of these "First Europeans" is found only in residual levels across Europe. The situation in eastern Asia is less clear in the archaeology, but I suspect similar dynamics are at play. Why? There is a strange result out of evolutionary genomics where the divergence between Europeans and East Asians post-dates the first settlement of these regions by moderns by 10-20 thousand years (depending on the result). My initial skepticism was toward the methods, because they were out of sync with archaeology. Now I suspect that there are two explanations. First, gene flow between these two regions is decreasing genetic distance, and the inferred time since the last common ancestor. Second, the first settlers may have had only a marginal genetic impact, in which case the archaeology remains valid, but not as relevant to the genetics.