Exactly. Some things just can’t be studied as part of a double blind RCT. For example, see: Parachute use to prevent death and major trauma related to gravitational challenge: systematic review of randomised controlled trials

The perception that parachutes are a successful intervention is based largely on anecdotal evidence. Observational data have shown that their use is associated with morbidity and mortality, due to both failure of the intervention and iatrogenic complications…

The paper is funny, but the authors are making a serious point. RCTs are great when they’re possible, but just because they’re not possible doesn’t mean we can’t gather strong evidence anyway.