Most of this effect is survivor bias. The people that don't have a problem - regardless of their past or future mental health - don't attract as much attention.

"Reefer Madness" is inconclusive [1] (so far as agitating an underlying psychosis). Many proponents would regard this as a affirmative result, however it does go both ways.

> Is there proof that they have a tendency of triggering "latent" conditions, instead of generating the conditions?

Personally, I've had panic attacks (originating from paranoia) bought on by historical heavy use - the correlation isn't subtle at all. Did my paranoia drive me to the initial peace associated with the drug? I'd have a hard time arguing otherwise - it really, really helped at first. My life was altered positively days after the previous dosage. If I take it now the situation becomes unbearable in less than 5 seconds. Anecdote disclaimer, as well as not being the same psychoactive drug (but it is psychoactive).

[1]: https://jamanetwork.com/journals/jamapsychiatry/fullarticle/...

They definitely can potential prodromal cases that might not otherwise progress to a full blown case of mental illness and do so in a way that exacerbates it. Many cases of schizophrenia in the prodromal phase don't necessarily develop into a clinical case so it isn't simply isn't a matter of psychedelic use being net neutral since they would have gotten it anyways

Psychosis is literally damaging and the earlier and more violently one is subjected to it the worse their lifetime prognosis will be.

Lifetime prevalence remains the same.