I mean, it's not true in general if you actually read the article.

> Whilst there are some minor indications that the hypothesis could be correct (eg. many of the absolute worst restaurants being some of the closest) the correlation is simply too weak.