Thanks for that article, it confirms what I'd long suspected. The bit about traffic cameras is a good one, and the very same problem of trying to judge health supplements efficacy without RCTs.