When I was 5150'd, I would have been unable to engage with, much less benefit from any other treament method until I was stabilized on an effective combination of meds. At least effective enough to get me lucid and coherent. (Hashing out cost, availability, side-effects and overall effecacy are topics worthy of 100 PhD dissertations.)
I'll also spare a thought for novel (and controversial) psych treatments like ECT (still the "nuclear" option reserved for when everything else fails), spravato (esketamine, an enantiomer of what some might know by the street name special-k), TMS, psilocybin, LSD, even Ayahuasca. Of that list I've tried psilocybin, but only recreationally. (Good times, would repeat.)
My favorite example of how f'ed up the pharmaceutical industry is is pointing to diphenhydramine. It's commonly known as Benadryl and used otc as an antihystamine with the prominently listed side effect of drowsiness. The same dose of the same chemical is sold separately as Nytol, a sleep aid which curiously lacks any mention of the side effect (treats allergy symptoms). They don't know what a substance will or won't do when they engineer it--they frankenstein molecules until they come up with something patentable and then put it through trials to see if it kills people and, if not, how best to market it based on its outwardly observable effects. Better than nothing, but emblematic of how much pharmacology relies on brute force and blind luck to achieve its desired aims. Even then it continues to be profit driven--an inherent conflict of interest in any field--and governed by the same "good enough, shove it on the market" approach employed by the software (particularly videogame) industry.
"Is it fun yet? Yeah? Forget the rest of the bugs, we've got a release date to meet. We can issue a critical patch later if necessary."
Besides, there's no larger pool of defacto QA interns happy to work for free than the pool of customers using whatever half-baked product gets cooked up and served to them.