That seems a bit of copout though. Yes, sugar pill will not have a as high placebo effect but it is still useful as setting a lower bound on the placebo effect.
Wouldn't it complicate things by introducing a new variable to consider, i.e. how gullible is each participant / how affected were they by the placebo? For what it's worth, it seems more useful to not have that variable. That way, if the effect was from the psilocybin acting as placebo, at least you know they were all affected equally. There were too few participants too. Having a control group with a placebo would make the experimental sample even smaller.
