To be fair, isn't a major factor (if not
the major factor, judging by MSAR...) in gaining top tier interviews the MCAT, and top undergrads tend to have higher MCAT scores? Someone from MIT told me that their average MCAT score for accepted students there was something like a 518. In contrast,
a top state school like UCLA has an accepted average of 514 (even though there's much greater weeding out at UCLA). Probabilistically, there's just a higher chance of top undergrads having students who obtain higher MCAT scores, so you'll see more of them on the interview trail if you interview at top tier medical schools.
From talking with admissions officers, it never really seemed like which college you attended mattered as much as people like to think. They always mentioned that students from these schools tend to have higher MCAT scores or have access to excellent research and so forth. I think jonnytest mentioned that it was a "mild" bump but in the grand scheme when the committee goes over an application, either during the interview screening or post-interview, I just can't imagine it meaning very much when the committee is more focused on experiences and interview performance than by the name of where you went to undergrad.