Today we should assume that it's harmless to get the placebo.

Only if a trial shows extraordinary consistent benefit without drawbacks would they stop it because the placebo group is harmed for lack of access to the treatment.

I used to share this sentiment for Parkinsons and other currently-intractable conditions. I still do sometimes, depending on the circumstances of the patient in question.

But the thing that stopped me generalising this was seeing second-hand that trialing a new treatment can be extraordinarily dangerous (and/or fatal).

Imagine being in the experiment group and the untested treatment being a catastrophic failure.

Imagine symptoms improving being in the placebo group?

What are the statistics on that?

Being part of a rigorous process by which a cure may be found? Sign me up.

You're thinking that even a gamble on untested / unverified treatment might improve their lives; it might, but it might also do nothing, make it work, or worst case kill them.

But I get it. My partner has an untreatable condition (like EDS, but without the genetic markers); they'd jump on any kind of hope, even experimental at this point.