None of us have the stats to prove anything conclusively.

And this is the problem. If the article is going to make such big claims with such small evidence, then it seems worth evaluating some more obvious solutions first.