I forget which interview I was listening to about doing LSD research but I remember the gist of the quote: "We tried doing double-blind studies, but 15 minutes into the experiment we realized it was pointless. The people who were given LSD... they knew they'd been given the real stuff, the scientists knew very obviously who'd been given the real stuff... There's no way to give someone LSD and not have everyone in the room know what's going on."
That's a very interesting problem. I wonder if one solution to the 'control group' is for the control to be an alternative intoxicant (i.e. not another psychedelic).