The body of stringent studies on non-pharmaceutical medicine is immense. Most of it has taken place outside of the U.S. (the most drug-dominated country), in Europe, India, China, and elsewhere. Until recently I suppose someone who was inclined to be dismissive could write off the India and China studies as being done by primitives, but it's not the case now, and it wasn't the case then.
It has nothing to do with who is doing the studies and it has nothing to do with what is being studied. Science isn't a dogma or set of information, it's a
process which works exactly the same on anything being studied, whether it's some pharma company drug or a traditional Chinese herbal remedy.
It has to do with how the well double blinded and placebo controlled test is done, how many people are included in it, and whether it can be replicated by someone else using the same procedure. For pharma companies trying to develop a drug, this is almost always done properly and well in excess of 95% of potential drugs are rejected because they fail these straightforward tests. This is why many drugs are so expensive, all those failures are expensive.
And when they fudge the data, we get crap like Fen-Phen. You will notice that faking or ignoring science led to that problem and doing the real science identified the problem. That's the point, the scientific process is self-correcting. You just need to do the tests to prove it.
The exact same process is done on TCM, acupuncture, herbal remedies, homeopathy, etc. and they fail these scientific tests. It's really a very simple thing: is the effect repeatable/dependable and does it have an actual measured effect greater than placebo? Also if it's putative drug (ie: herbals) does it have a dose-response curve? Ie: if you give a greater dose, does it have a greater measured effect?
If there's no effect to measure, then what's the point? The alternative remedies which fail these tests are what get marginalized by mainstream medicine. Because they don't work. So all we get is special pleading that maybe the next time or the time after that it'll really work.