You cherry-picked that example. (Plus took it out of context—the quote talks first about food and medicine.) But even granting the point completely, the presence of such outcomes in a scatterplot says nothing about whether the approach is beneficial or not on the whole.

I find it exciting that this is at least partially amenable to objective experimentation. Why not try it and see? It takes a pretty rigid ideologue to not be open even to that much. (Not talking about you there, but about "they'll only spend it all on drugs" and so on.)