Human evolution is a field that's been evolving very fast lately. Even people who graduated only a few years ago are wildly out of date if they haven't kept up.
That said, classical multiregionalism is very, very dead and has been for awhile. Essentially what you're asking about has started to come up lately as a sort of in-Africa multiregionalism. This is a nightmare to model mathematically and so people just couldn't until recently when new fossils made it pretty hard to explain things any other way.
It hasn't fully emerged yet and many people don't really deal with it. Take for instance, this paper. One of the fundamental assumptions is that there was a singular set (or other time limited) of introgression events between archaic and human populations. That's a reasonable assumption in older models (and dramatically simplifies things), but it's possibly violated in an African multiregionalism model.
They claim it's not a problem because of symmetry, a point I'll admit I don't fully understand their explanation for in the supplementary material.
Nah, it had nothing to do with PC. Back then, PC was about labels: "disabled" (they tried to get "differently enabled" to stick!) vs. "handicapped," "little people" vs. "midgets" and stuff like that. You were even supposed to say "queer" instead of "gay," and some were just starting to use "gender" in non-grammatical contexts like referring to a person's "gender" (which was look-it-up-in-a-textbook incorrect back then).
Out of Africa feels less PC to me, but I won't pretend to understand the rationale behind what is and isn't considered PC a lot of the time. But I agree with GP: we have relatively few DNA samples and try to draw pretty big generalizations between them. Ituitively, I would expect patterns of migration in and out of Africa and all the continents to be far more complex than 1-time events from which entire populations then developed complete independently.
edit: On a related note, my siblings' DNA test results say that they're something like 4% Native American, yet we have very reliable documentation of pure British genealogy back on all lines almost all the way back to the 1500s. Very unlikely to actual have a modern link. I'm sure the companies are likely overplaying the similarity more than anthropologists would, but am I to conclude that I have a closer link to Native Americans than other random samples from Europe?
The DNA tests just look at haplogroups and mtDNA lineages etc. to make _very_ broad generalizations about what population set some of your ancestors _may_ have belonged to.
It's more likely that at some point in the last 15,000 years someone in your lineage had a child with someone with some Siberian background way far back.
I'm not stating any claim to truth here, but which one sounds more pc? Caucasoids diverged from negroids 100k years ago. Or, Caucasoids diverged from negroids 2 million years ago.
You can look at the data yourself, it's not hidden.
Also, the classical meanings of terminology like "negroid" or "caucasoid" doesn't really map to genetics. My personal experience is that they're almost exclusively used by people who aren't familiar with modern understandings of human evolution. There are a lot of cranks talking about it, so it's often best to avoid archaic terminology that might get you mistakenly grouped with them.
I think accusing something of being PC implies it's not actually correct. So if you're saying one of these is more PC, that's a claim that it's less factual. Otherwise gravity is PC, 1 + 1 = 2 is PC.
Perhaps they dumbed down the complexity for a bunch of freshmen, but I remember wondering then why some combination of the two wasn't possible.
It certainly looks like a mix of the two canonical models matches the data that is being revealed through genetics.