Nope, your response is based on misunderstanding. Point by point:
Wrong: I don’t believe that “any classification of people into different racial groups is invalid”. It’s possible that somebody may at some point come up with a classification scheme (say, skin lightness times follicle roundness times earlobe length, or whatever) that does use phenotype similarity as a scientifically reliable proxy for genetic kinship.
But at present, nobody has successfully come up with such a valid classification. It’s not just that attempts to treat currently defined racial groups as somehow representing genetic populations are unproven: it’s that they’re often factually, demonstrably wrong.
If you say “A black Australian Aborigine and a black sub-Saharan African are the same race, and therefore we can assume that they are genetically more closely related to each other than to a white European”, the problem with that is not that you’re being “politically incorrect” or something, but that you are literally, provably, scientifically WRONG.
DNA studies make it clear that those two “members of the black race” are NOT more closely related genetically to each other than to white people. And that makes complete hash of the concept of race as some kind of cross-cultural global category that reliably indicates genetic kinship.
Likewise, if you say “The light-skinned heavy-bearded Ainu Japanese are genetically more like Caucasians than darker-skinned smooth-faced East-Asian-looking Japanese”, you are just plain WRONG.
So if racial classification based on phenotype characteristics can’t reliably predict genetic relationships, then what’s the use of it?
Nope, it’s because studies on different populations within “racial groups”, like the ones on Aborigines and Ainu that I linked to above, have scientifically shown that the “racial group” classification is not reliable.
Now, that’s certainly not to say that some populations within a given “racial group” aren’t closely related. I’ve got nothing at all against studying genetic differences between genetic populations that are correctly identified as genetic populations.
But throwing a bunch of demonstrably genetically different populations into a single “look-alike” category called “the black race”, or “the Caucasian race” or “the Oriental race” or whatever, and claiming that their membership in the “same race” should be taken to imply genetic similarity, is simply horseshit.
It would be like a stranger coming up to the three Rover brothers whom I hypothesized in my previous post and saying “Well, clearly the blond Rand and his blond brother Ron are more closely related to each other than to this darker guy Milton”. Sure, it’s not unreasonable as a superficial guess, but you already know, and can scientifically prove, that it’s simply WRONG!
If the three of you are all full brothers, then you are not more closely related to one of them than to the other, even if one of them happens to look more like you. And anybody trying to draw any conclusions about the Rover family based on such a mistaken assumption would merely be a horseshit-generating horse’s ass.
And that’s what people are who attempt to draw conclusions about genetic differences between cross-cultural trans-ethnic phenotype-defined racial categories which contain population groups that have been proven to be no more genetically similar to other populations within the same “race” than to populations outside it.
The reason they’re invalid is not that they’re using racial categories per se, but because they’re jumping to conclusions about the genetic kinship of population groups within the same racial category that have been scientifically demonstrated to be—what’s that word? oh yeah—WRONG.