I'm now more sure you're riding a hobby-horse. I'd better explain why.

Well, speaking empirically, the phrase "race is a social construct" is pretty often followed by "therefore all races are the same in all important ways". [...]

It's not that it is a personal hobby-horse, it's just that I have some experience in watching similar conversations develop.

Me too. Which is probably why I had a hunch that, from the start, you pattern matched Elund to the kind of person who says things like "all races are the same in all important ways" — because you'd observed such people before — in spite of Elund not having said that. That hunch now seems to be confirmed.

That pattern matching would make sense to me if, say, in the context of an argument about race & IQ, Elund had started insisting "race is socially constructed so racial IQ differences can't exist haha I win!" as a desperate gimmick to shut down the argument. But the context wasn't a fraught debate like that; Elund's "Race is a social construct anyway" was an aside to explain why they were content with someone treating "mixed race" or "Hispanic" as racial categories, which doesn't sound like a mind-killed person invoking "uh uh uh it's a social construct!" to evade an argument.

So the way you responded to Elund (asking a pointed but not especially relevant question about what doctors think; intimating that Elund was doing an intellectually dishonest post-modernist two-step; asking a question which falsely implied Elund said race wasn't a useful concept; and dragging IQ (hitherto unmentioned) into the conversation) didn't seem consistent with a dispassionate correction. It looked a lot more like taking a hobby-horse out for a canter. Reviewing the argument, I'm not sure I could come up with any empirical question about race where the two of you would disagree on the answer!

"Social construct" implies an arbitrary choice -- our society decided to split humanity into races this way, but another society might do it in an entirely different way and all such ways are equally valid, which is to say, there are no underlying "real" differences.

Supposing someone wanted to split humanity into arbitrary races based on actual genetics (which is not how the concept of race originally started because genetics wasn't known at the time), it would make sense for most races to be African, since Africa has far more human genetic diversity than all the other continents combined do. The reason races are delineated the way they are now is due to social reasons. (It could possibly make sense when you consider the phenotype though, but due to the outgroup homogeneity bias, I have some doubts.)

Still, regardless of where you set the boundaries between races, there will be average biological differences between them (provided you don't do something biologically ridiculous like classifying whites and Asians as the same race but then classifying their half-white/half-Asian children as a different race).