If you truly belief in rationalism and don't engage in it to signal status, I see no reason to use another standard for judging whether homeopathy is true than for judging whether PUA works.

Aww, I respect you as a person too! (What were you trying to accomplish with this comment?)

As you point out, which control you pick is significant, but my point is that what test you pick is significant too. Let's talk about basketball: you can try and determine how good players are by their free throw percentage, or you can try and determine how good players are by their average points scored per game. You're suggesting the analog of the first, which seems ludicrous because it ignores many critical skills. If someone is interested primarily in getting laid, it seems that the number they care about is mean time between lays, not percentage success on approaches.

I won't comment much about your homeopathy example, except to say that even if one considers it relevant it undermines your position. Homeopathy is better than both nothing and harmful treatments (my impression is most people come to PUA from not trying at all or trying ineffectively). Generally, for any homeopathic treatment you could take there is a superior mainstream treatment, but for some no treatment is more effective than placebo (and so you're just making the decision of whether or not to pay for the benefits of placebo). Likewise, even if the only benefit of PUA is increased confidence, you have to trick yourself into that confidence somehow- and so if PUA boosts confidence PUA increases your chances, even though it did it indirectly.