Tajima's D test applied to tandemly repeated genes

Question

[jaer.wees]

Tajima's D test is normally used to test for evolutive contraints on a DNA sequence by analyzing population samples. In the following article, it is used to test paralogous sequences of a large array of repeated genes. There are no population variants analyzed. I wonder wether such an extrapolation of the Tajima's D test is correct.

PMID: 24945811

Thanks for advices!