One problem with the kind of markers mentioned in the cited article is that they are not subject to selection pressure and not likely to be correlated with any phenotype of interest.
The markers simply prove that race is, indeed, a definable category. This makes the question "is group A different from group B in way X?" meaningful.
It proves nothing about the answer, nor is it intended to.
For comparison, a less meaningful question is "are biologically NY people smarter than biologically NJ people?" The question doesn't even make sense since you can't differentiate between NY and NJ people biologically.
As for whether the markers are likely or not to be correlated with interesting phenotypes, I'll wait to see data before guessing about what is likely.
