This would only make a difference if P(low dietary vitd) or P(low solar intake) is small (and assuming dietary values/solar values are independent of having a genetic disposition to lower vitd levels and independent of COVID prevalence), which I don't think is true. What I think may be true and may not be compensated for by the study is the fact that individuals with genetic disposition for lower vitd levels may be supplementing (through either greater sunlight exposure or dietary supplements) to bring up their vitd levels, even before COVID came onto the world stage.

> I don't think it's an honest assessment

Of what though? This is just a study trying to prove a correlation with VitD deficiency and COVID prevalence, of which they couldn't find any. Any stronger conclusion is out of the scope of the paper.