It turns out these are basically the only conditions for which medicinal use is roundly supported by the data. Everything else is basically "Well, weed makes lots of people feel good, couldn't hurt to use it for [pain/nausea/anxiety/whatedver]."

True as it is, the limited data is mostly due to criminalization in the US preventing comprehensive analysis. This is especially true for the less severe conditions (of which you mentioned several) where well studied alternative remedies are effective and plentiful. The severity of suffering assosiated with epilepsy, the limited effectiveness and heavy side effects of previously available treatments and the peculiar effectiveness of cannibus eventually resulted in conscientious objection and civil disobedience which resulted in much more of that data becoming available.

I'm not disagreeing, just noting that opium made lots of people feel good long before the why was well understood (and well capitalized...). I look forward to reading more about it years from now, once we get past the Rumspringa phase of gummy bear concentrates at the gas station and drizzled over you're starbucks latte :-<