I guess it depends on how reliable you think PSMC methods are as they approach this time frame. The data looks smooth to me out to around 1.5 MYA or so, plus or minus, but the method loses its power as you go back further and further.
Yes, the main issue is that at 1-2MYA, human (or hominin) behaviour is nothing like what we consider normative for modern humans.
For my part, I think the whole exercise to find an ancient A & E is driven by concordism. I’d rather let Genesis be Genesis on its own terms.
*Note as well that I am not an official BioLogos voice on these matters - this is my own personal opinion, lest anyone be confused. I think approaches that have Adam and Eve within a cultural context that fits Genesis (i.e. around 6-10KYA) are much better ways to go if one is interested in a model that has a historical Adam and Eve.