I tell my friend, "I have a lucky rabbit's foot."
She asks, "Lucky, huh? Does it help with gambling?"
"Yes, it should help while gambling — I don't see why not."
What I just failed to realize in that conversation is that I literally just said that if I use the rabbit's foot, I will experience a higher rate of favorable outcomes. I can't wriggle out of the fact that I just said that — "Rabbit's foot helping with gambling" is literally the same thing as "Rabbit's foot improving probability."
Yet many people think of that second claim (and testing it) as though it takes place in some other world: Planet Empirical Science, where everything is made out of test tubes and lab coats. During their trips to Planet Science, both scientists and laymen simply visit and extract the occasional bizarre idea, like "quantum thermodynamics", then import it to here in the Real World, which is made out of Everyday Experience. I mean, you don't actually believe that physicists "observe" all those super-tiny particles, right? That's like saying medieval theologians "observed" angels dancing on pins, right? In both cases, the claims are so "out there" that they must be made up, and therefore, why, we have the license to use those claims however we please.
So it is with homeopathic claims: they don't just say that it makes you feel better, but that it outperforms conventional medicine — and therefore, although they don't say it aloud, that it outperforms placebo. Additionally, if "water memory" were a "quantum" phenomenon, it would be empirically observable as such. That second part is perhaps even harder to grasp — I know I don't fully grasp it (perhaps the resources to actually test the "quantum water" claim would be greater than those devoted to the LHC, but in any case, it would still be ultimately testable).