Agreed, but I'm not sure the solution is upping the statistical standards so much as educating researchers about what the statistical standards actually mean, how to use them, and how to construct a valid, repeatable experiment.