I've been doing a lot of research about this as I choose between a school with a mostly lecture-based curriculum and one with a PBL curriculum. I'm leaning towards the PBL school, though the curriculum is only one reason. Anyway, I've been talking to a lot of older students on the phone–I've been tracking them down at both schools–and what they all say is that in academics, age isn't an issue. At some schools, age is the major determining factor in social groups; in others, not even that is determined by age. This seems to vary even class by class within schools. Whether outside study groups are harder to form as an OPM is probably similar in either kind of curriculum, depending more on the cliquishness of the school than the curriculum itself. In several schools, I've actually found that the students in their late twenties and early thirties do cluster as a group, but that the older students (late thirties and up) sometimes are just as happy to hang with the twenty-two year olds, and end up in study groups/social groups with them instead of with the slightly less young students. I don't know why this is exactly, but I've heard this about the oldest students at several different schools. Is this because the much older students don't have any investment in their own ideas of their own maturity, and so can just enjoy the company of people very much different from them; whereas the somewhat older students are more interested in proving to themselves that they are quite different than the younger ones? Just a theory–we'll see how it plays out. (I'm 32.)
It also seems to me that small degrees of variation in performance in a given class are not going to be as big a deal in many schools' preclinical years, esp. most of the PBL schools as far as I know–b/c they're pass/fail. So, assuming that I go to a school with pass/fail preclinicals, I expect to be competing less against others and more against myself and my own expectations. It's not like pre-med where the difference between an 88 and a 93 on a project seems like a huge deal, and can be influenced by whether your lab partner is a problem or a help.