I disagree, power isn't a requirement for discrimination, you have wrong priors.

In fact making that a premise leads to quite a few assumptions that are clearly wrong that falsification is trivial, even if you could nail any sensible quantification of power.