It's not the ideal case, but a sample size of n=2 people could be enough to conduct a preliminary study. I'm thinking it'd span over a reasonably long period of time, maybe 3 months or so? Space the doses out with 3-4 weeks in between, and give 3-4 different trials. The dose, ROA, and environment should remain constant. It should be set up so each participant recieves both meh and magic over the duration of the study, to ensure that both types of M get a fair representation. Of course, more data is better. I can't imagine it would be that terribly difficult to get some people to volunteer to take MDMA and hang out in a beanbag chair watching 2001 or something.
Study participants should be given an overnight fast and a light breakfast, a urine drug screen (no sense in having interferences from other drugs), measurement of vitals, then followed by adminsitration of the test drug at e.g. 10AM. The test is conducted in the same location, which ideally should be comparable to the setting of other psychedelic therapy settings. Beanbag chairs, lava lamps, art on the walls, y;know - a chillout space, basically. Provide activities to keep the MDMA-addled subjects entertained for 6-8h. During the experience, data could be collected to validate the effects profile(s), either psychometrics, biometrics, or just narrative experiences. Upon conclusion of the experience the participants are once again checked and then released to go home.
Then three weeks later, we do it all over again, with a new set of randomized capsules. Repeat another two times or so, or as needed to provide the stastical significance. (I don't see a need to include a placebo, but that could be given too, if you feel like it.)
Yes, this is a very involved study. Probably outside of what the average BLer is capable of organizing. (Anyone wanna talk to MAPS?) I think that it'd be worth the investment in time and effort. Think of the prestige, being able to say you knew what the difference between Meh and MDMA was, and have the data to prove it exactly! Even if it provides a null result, you'd be giving people free empathogens, and furthering science as you do it.