You can't just say "X is difficult to encapsulate" with studies. This is a fully general counter argument to any evidence you don't like.

That's not what I'm saying. I'm just saying that even if a small percentage of the population took that advice as definitive and executed over it, all latent small probabilities would be realized and the mistakes would be immediately manifested. Stuff like that isn't being difficult as any pharmaceutical company or any medical company has to know for sure that their studies/devices are correct for health effects. In fact in statistical decision theory texts pharmaceutical examples are standard.

tl;dr this would be a big lawsuit of some sorts