Except there are sources (that I'm not going to go find, because I believe I've cited them before to you, and I'm not really going to call my house and find out what medical journal they're in) where homeopathic remedies preform better than placebos in side by side testing. There are also plenty of journals that say they preform worse than placebos.
To point out, if a treatment is completely useless then yeah, that is what we would expect to see! There will be some unknown number of people getting better from some unknown factor, and some unknown number of people getting worse from some unknown factor. We call this noise in in statistics, because it doesn't give us information about how good the treatment is.
If you want to pay money for a treatment, chances are you want to see some positive effect from the medicine, which means at the absolute lowest bound you want to see more people getting well than getting worse and having enough of the former that it swamps out the noise.
The results being inconclusive is a conclusive indicator that the treatment is ineffective.
--
You're messing with me! You're messing with me, aren't you!?
You're making fun of me, aren't you!? Aren't you!? You definitely are! I'll murder you!