I don't see any misuse of statistics in the article. First raw data shows that a strong correlation does exist. Then, before interpreting that in terms of probabilities (with due care),
a plausible explanation for causality is offered from statements that appeared in the press. Correlation seems far from being due to chance.