The categorization is superficial because there is no use of it
The concept/construct which I call "race" is more or less equivalent to the concept/construct which population geneticists call "biogeographic ancestry" group. If there is no use for this, why is it so frequently employed in population genetics and why can I find hundreds of recent admixture mapping studies which examine the association between ancestry with respect to major "biogeographic" groups and medical related outcomes?
Races will eventually intermix and cease to exist as a concept sometime in the distant future.
In the unlikely event of perfect human panmixia, how would the concept/construct per se cease to exist? The concept/construct allows one to express the proposition that "there are presently no human races"; if it, as an epistemic class, ceased to exist, the proposition would be rendered meaningless, equivalent to ""there are presently no agfhddsf", no?
Do you mean, instead: "In the future there will be no distinguishable human biogeographic ancestry groups"? In my assessment, synthetic biology will allow for accelerated divergence and interstellar travel will impose substantial reproductive barriers, so I expect, in the long run, speciation. Do you disagree? If so, why?
Note that I have made all of these points in my paper. Maybe try reading it and after formulating a critique.