Looks like you've stumbled into a better causation/correlation hypothesis than the one in the article.