For a community that prides itself on critical thinking, I'm always surprised to see HN lap this sort of pseudoscientific witch-doctor stuff up. This poorly-controlled, N=1 experiment tells you nothing, not even about the author. There's absolutely no reason to consider these novice self-experiments when professional scientific experiments are available (unless you're hunting for a specific result).
Also, N=1 experiments can absolutely be interesting and give us ideas for further study, even if they don't say anything about a population.