I think it's impossible to do a double blind experiment, but you can still test it to compare vs placebo and vs standard medicines and vs another psychedelic.
The idea to compare to a placebo is that there are a lot of variables that you can't control. For example, most studies have some eligibility criteria like age, so you need to know the baseline reaction of your population.
Perhaps the weakly meeting to check the general state of the subjects is enough to change the outcome. If you don't compare against a control group with almost the same treatment it's possible that the effect you are measuring is not due to the psychedelic but to the other details.
This is a different argument about the difficulty of testing psychedelics; it emphasizes a problem about the effects of cultural context (and expectations?) for drug experiences. So it might be hard for an experimenter (or a clinician!) to reproduce factors in someone's cultural background, context, or expectations that might have a big effect on the patient's experience.
On the other hand, with the mental health profession paying a greater amount of attention to "culture-bound syndromes" (mental illnesses that occur in some cultures but not others) and to cultural aspects of the experience of mental illness, maybe this problem also affects tests of things other than psychedelics.
