I clearly stated beginning research in animals. Your metric for determining if in-vivo experiments should be done on animals is interesting. Based on your notion, you should only perform these experiments if you think they will actually work. The problem here is that the metabolic processes of mammals for instance is so complex that we often times have no idea what will be effective and what will not. Oxidizers is a classic case. It is commonly understood that anti-oxidants are good for the body. But you can find research that states that too many anti-oxidants can be bad b/c it inhibits some amount of oxidation which can be good (b/c it kills pathogens).

I am simply advocating investigation based on research, realizing that we actually know very little about oxidation vs anti-oxidation within complex organisms. If we only ever think "inside the box" with what we "think is probable" we will almost certainly never discover many of the mysteries in our biological world.