Someone else compiled the basic stats involved (which I didn't cite due to possible blogger gripes), but didn't point out the correlation, which I noticed on my own. The middle 10 teams work out as follows:
If you add in the middle teams, the correlation becomes somewhat weaker, but is still statistically significant (p = 0.026, Pearson). So, yeah, there's a pretty decent argument that the better teams are favored with fewer back-to-backs.