It's a social construct in the way that all categories are. The question is whether or not a category is useful. I'm saying there are genetic differences between groups, but our structuring of these groups is pretty flexible. There are genetic differences between "continental groups" but there is also a good amount of genetic diversity within sub-Saharan Africa populations because those people did not go through the genetic bottleneck that Eurasians did.

Our understanding of race have been historically contingent. Both the racial essentialism of the colonial period and the "we are all the same under the skin" anti-racism of post-WW2 were based more on political ideologies than reality.

If you tell me you have a test that can distinguish a Japanese person from a Finn, I'd say sure. If you told me you had one that could distinguish a French person from a German, I'd laugh.