yeah some similarities did converge, like different skin and hair per environmental factors (always on a gradient though) or adapting larger lungs if you are at higher elevations... but we already know that we look different. that's the whole problem, that we can't get over how we look different. unfortunately physical appearance tends to be rather important on the primate line. primates are very visual creatures.
i think you could *obviously* get different behavior traits through evolution for different groups of people... but humans haven't been that controlled or isolated in their groups. humans tend to wander and trade off with different groups. but i'll use the dog breeds argument... not only do dog breeds look very different but there are behavioral differences... for instance certain characteristics are associated with herding dogs vs. with retrievers. herding dogs need to keep track of an entire flock of grazing animals and follow complex commands involving various signals or words. is it possible that the average border collie can learn more words/signals, is more likely to herd small animals and children by instinct, and is more likely to always being vigilant of an entire group and making sure they're all together? sure, i think that's reasonable. because along breeding lines there are certain behavioral traits breeders were looking for (the dog needs to be good at performing a particular function), so they were probably more likely to breed the most exceptional seeming dogs (although of course training is a *huge* factor in if the dog was exceptional... ahem), hoping to keep getting the desired traits. of course there are numerous issues with this since it *is* so dependent on training and since it's hit or miss in breeding (it's way easier i bet to select for physical features than something abstract-ish like "behavioral tendencies" - with the latter you really are just guessing). some individual dogs of breeds not even supposed to have anything with herding could turn out to be better herding dogs than some individual dogs of a herding breed. and of course how do you even define "better" in a way that is scientifically valid? also, breed your herding dog with a mutt and boom, the offspring just looks like a bunch of dogs. how quickly you can unravel your "breed" sort of demonstrates how possibly insignificant the little changed you have made to 'dog' are.
i don't think anything in human evolution has been so controlled and fixed as dog breeding, save maybe some sick eugenics experiment forced on people. many scientists argue that you can't even find worthwhile differences in personality traits or cognitive abilities or whatever between dog breeds - and that's with actually *trying* to breed for very specific traits.
i could buy that humans have different races if the world consisted of 4 continents all separated from one another by ocean with 4 separate human populations on them, isolated from one another for at least 1 million years with very different environmental conditions on each. if you bring the 4 groups back together after 1 million years, you might find they are still the same species but that there are considerable differences... maybe you could measure some of them. because you know how long they were isolated from one another; and because it is "controlled," you can probably be somewhat safe in assuming that a lot of the observed differences are actually significant. however, you would *still* have to account for culture and conditions and not confuse that with actual biological differences. and you might actually even still not be able to find distinct differences in personality or cognitive abilities of statistical significance--like consider horses and donkeys, two separate species that imo are largely the same lol (despite who know how many years of evolution apart they really don't seem to have diverged much from one another). or consider a species that lives on multiple continents like the grey wolf... are there significant differences (they've been separated in different environmental conditions) to write home about? (you might automatically think there aren't because they all look the same
)
anyway, the point is though that human evolution has not been nice and clear cut like this. there has been a lot of overlap between groups of people, we're at most 200,000 years old as a species... we don't have the proper controlled isolation or time or controlled vector (i.e. relentlessly breading in or out *specific* traits each generation) for there to be anything significant. that is to say you cannot find a way to break down humanity into meaningful categories based on any set of characteristics (subtracting out cultural factors) that tie to their genetic heritage. so you can't actually have a "race."
i think the other plausible seeming thing is that you can get morphological differences out of evolution pretty quickly and easily. it makes sense. i mean if a species can't get a needed morphological adaptation for its environment fast enough, it'll go extinct. the things on the surface of us (the literal physical surface) are kind of our first line of defense from the environment. the muscles and interior physical shapes like that of a blood cell may be our second line of defense. these things really seem much simpler than "behavior."