First of all that could be a causality ≠ correlation, but more importantly. Even if it might be the biggest predictor, it might not be the biggest lever.

Still interesting train of thought that I might need, but the article is exaggerating the evidence and not adding the right disclaimers.