Too much depends to judge the efficacy of the MMI itself (does the interview solely decide to accept the candidate?). But a bad MMI implementation (and there is wide variance from the other threads posted) is just as bad as a bad traditional interview, in the end it's the people running the show that really count, in both cases.
From what I've seen the MMI is an attempt to create a process to make up for the shortcomings of bad interviewers. But in the process it hamstrings a good interviewer, who presumably has experience with introverts as well as extroverts and knows not all that shines is gold.
MMI is great for schools with bad interviewers and tries to "spread the risk," but on the other hand, it is terrible for schools with an already great set of interviewers in place. Granted, it is difficult time consuming and resource intensive to build a good set of interviewers.
But right now, we are all debating preference because we dont know how it fits together in the end and are just expressing our reactions to our small set of app/interview experiences. I agree we should move the MMI debate to another thread like this one:
http://forums.studentdoctor.net/threads/how-to-prep-for-mmi-interview.1041399/