If I find a meta analysis that agrees with your claim, but has obvious methodological flaws you could simply say "you picked the wrong one" or "that wasn't one of the limited domains of application".

I did as you asked anyway, in my pubmed search 5 out of the top 10 results are by the same author and are published in the journal "Homeopathy". I have no doubt there are many positive studies in that journal, just as there are many true sasquatch sightings in "Sasquatch Magazine".