Well, at least in the case of resveratrol, there is ample evidence to suggest that it does indeed impart salutary health benefits. I won't reproduce my grant application introduction, but resveratrol has been shown to reduce the incidence of cancer, heart disease and neurodegenerative diseases in animal models, admittedly at quite high doses. David Sinclair of Harvard has done some of the most interesting studies, showing that high doses of resveratrol increase the lifespan in certain model organisms (roundworms and zebrafish) though not, alas, in mice. However, Sinclair did show that high doses of resveratrol were able to counteract the health consequences of obesity in a strain of genetically obese mice, reducing the incidence of heart disease and diabetes to levels seen in a control group of mice. The problem for humans is that, to get comparable levels of resveratrol from wine we'd have to consume 200 bottles a day.
The study that gave rise to the notion of the "French Paradox" took into account dietary issues, too, showing that the wine drinkers consumed comparable levels of fat to the control group.
Mark Lipton