It often takes millions of dollars if not decades to prove a particular point of scientific evidence
This is sort of like saying that because an 8 year old can't do statistics they also can't add 2+2.
Your example is based on complexity. For instance, if you listen to any nutritional scientist, they will tell you that nutrition is a vastly complex subject/issue, and that we have barely scratched the surface on the mechanics of indivitual nututrients let alone chemical/nutritional interactions between individual components.
Can you hear a difference is a 2+2 kind of problem. It dosn't take millions of dollars and years of effort to tease the answer out of complex data or interactions. All you need to do is switch between two components without the listener knowing which is which and ask, "Did you hear a difference?" This is a question that can be asked a number of different ways to guard against biased language.
On the other hand, if you believe whatever component makes the difference, spend your money and be happy.
