> The danger is when/if people selectively choose studies

This, in itself, assumes that all studies / papers are equally as good and methodologically sound, which is not a true assumption.