You said the social construct of race produced in the past isn't supported by molecular methods (correct me if I got that wrong), do molecular methods show groups that could still be classified along racial lines if we ignore prior attempts? Seems to me we're just replacing 'race' with some other word based on newer, better methods. Course we're letting the science do it rather than the personal biases of past peoples, and thats a good thing.
You could identify groups of people that are genetically similar to each other and somewhat distinct from other groups. But I would say that these classifications would be along "racial lines" would be misleading at best, because the classification would depend on the algorithm. One algorithm would result in very different groups than another. and the number of groups would depend on how many groups you would want there to be. So method A would put you in one group out of 50, method B in a different group out of 1000 and method C would classify you as an outlier that does not belong to any group at all. That means, human race is an unscientific concept, because it is ill-defined and there is no obvious way to make up a definition.
You could force a definition by choosing one classification method and call that "race". You could then go ahead, subject every human to a genetic test and write that result in their passport. But why would you do that? What would be the benefit against the obvious drawbacks of having yet another set of dividing lines that creates conflicts between humans.
Obviously, whatever method you would choose, the result would be very different from the social construct of race. For example, the notion that there is such a thing as a "black race" is inherently racist, because biologically it is utter nonsense. There is always the danger of racists trying to shoehorn scientific facts into their prejudices and claim that there is a scientific basis for their racism.
I think the idea that there wasn't traffic between these populations is pretty ridiculous, the origins of flint-stone tools and the like can be pretty easily traced and we can see far-reaching connections from very early on. This all depends on what perspective on time we use, if we view it as homo sapiens versus homo erectus where it is about a million years between their respective spreads across the world it's obviously a bit different, but with homo sapiens there is a clear trade connection in just the early spread from east Africa to south Africa far, far earlier than the migration out of Africa (part of what makes Sapiens so advanced). Populations of Homo Sapiens were always connected in the times of the actual spread out of Africa while there were some distinct precursors some of which are linked in this thread like Homo Sapiens Idaltu (with Denisovans and Neanderthals being earlier, more "successful" verions of these precursors around 400k years ago
https://en.m.wikipedia.org/wiki/Timeline_of_human_evolution, this timeline should show approxomately how long it would take for Homo Sapiens to diverge, hundreds of thousands of years, not to mention Erectus lasted about 1.5 million years and diverged quite a bit while still being considered one species.).
How many non-marginal examples do we have of humans remaining actually isolated from each other for that sort of timeframe?
The only examples I can think of are indigenous Australians, who separated off about 60-70,000 years ago, and those of the Americas, who separated off somewhere between 16-40,000 years ago, depending on who you ask. And in neither case will you find consensus in favour of total isolation;
Maybe 10% of the world's population, at the very outside, have any degree of Amerindian ancestry. Less than 1% have any degree of Aboriginal Australian ancestry. So even if it were true, so far as it applies to these specific populations, what conclusions are we actually supposed to draw from that, as regards the other nine-tenths of the world's population?
There doesn't need to be total isolation for genetically distinct groups to emerge. Bottlenecks that limit genetic flow are probably enough for that. Even if population groups are still connected by a few traders, they would still evolve somewhat separately if the rate of genetic change exceeds the rate of genetic exchange. And there are enough geographic bottlenecks that limit exchange, like mountain ranges, deserts, or bodies of water. Sure, these can be crossed, but how often was this done? Usually, people didn't travel the world to find a partner. It is evident that the interconnections between humans weren't enough to fully mix the human genome to the point that local variations are greater than the global variation. For example, the African populations that were analyzed, carried significantly less contribution of Neanderthal genes than the analyzed European or Asian populations. So 40000 years of human connections weren't enough to carry those genes to those African populations.
The only way to avoid distinct groups would be to claim that the genome is a continuum to a degree that makes it impossible to place boundaries between groups. But there are enough geographic barriers that are inconvenient to cross that I believe it is possible to draw boundaries between genetic groups if you really want to. That doesn't mean it is a good idea to do it every time you can.