if it's working, it's working when you test it rigorously. one such big screening of traditional medicines happened in china somewhere in 70s: they collected 2k+ traditional malaria medicines, of which 350+ were tested on mice and found ONE new compound that had any activity and could be used safely in humans, so yes, there's plenty that didn't work

getting new active compounds is hard. most of the time it's easier to look up to nature, because some fungi or sponge had millions of years to get that one poison that wards off predators just right, and because of that it might be active in humans as well. whether is it useful clinically is another question. using things from traditional medicine is another filter on top of that, but far from perfect based on limited diagnosis available to ancients, limited to nonexistent disease mechanism understanding, limited plant availability, magical thinking, and more

and yes, there's plenty of things in traditional medicine market that are known not to work. realgar, shark fin, powdered rhino horn, nigella seeds, cow piss and many more do absolutely nothing at best. intricate magical systems behind their inclusion in medicine don't add any value or validity