Politically correct persons' hope that their labeling of race as a social construct will block any inferences made on the basis of race is unsound: chairs are social constructs -- not belonging to any science as a theoretical posit and apparently existing merely as somewhat arbitrarily gerrymandered concatenations of atoms -- yet we can still infer that being hit over the head with one will hurt.
So too with race: there aren't purely biological necessary and sufficient conditions for belonging to the class 'African American,' but we can still infer and generalize on the basis of the somewhat arbitrary set of properties that we gerrymander into our social construct, e.g., we can infer about persons born in the US with significant genetic lineage leading back to Africa with albedos between x and y that they often evince above average fried chicken loving behavior, or that they evince higher rates of diabetes.
Fitting Mr Wade's claims into the above construal of social constructs, he is simply saying that the genetic portions of our social constructs of race allow us to make inferences and generalizations about the objects denoted by such concepts, a perfectly reasonable claim. And a perfectly uninteresting one.