> Without controls and blind tests, I can understand how folks fool them selves into thinking something like that can work.

Fooling themselves is perfectly possible with controls and blind tests too..