You're not wrong, but those are definitely edge cases. You couldn't give placebo MDMA to a group of 50 people without the majority of them calling your bluff.
The problem then becomes the "nocebo" effect, wherein the patients who know they haven't been dosed with a powerful psychoactive substance assume that their treatment will be ineffective, and therefore stop giving it the required effort.
With therapy, it's important that the patient be onboard with and believe in the therapy. Giving people obvious placebos is a reliable way to lower expectations and instill doubt.
The problem then becomes the "nocebo" effect, wherein the patients who know they haven't been dosed with a powerful psychoactive substance assume that their treatment will be ineffective, and therefore stop giving it the required effort.
With therapy, it's important that the patient be onboard with and believe in the therapy. Giving people obvious placebos is a reliable way to lower expectations and instill doubt.