Boston definitely is not as fast as would be implied by the net downhill statistic. I would bet that if the course were perfectly pancake flat, we could've seen some 2:02's on Monday.
I once ran a road race where one mile was almost entirely with the prevailing winds. I couldn't feel it, because the wind was going about that same speed that I was, but when I got my splits that mile was 15 seconds faster than the one before and the one after. It's really basic physics. There is a reason for the rules that are in place - London and Berlin CANNOT have prevailing tailwinds the entire way because they loop. Like a track.
Ryan Hall ran a great race, as did Mutai, Mosop, and Geb Geb, but it was seriously aided. It's no different than Steve Scott and John Walker still being great milers, even though I can say that their downhill mile times shouldn't be considered PBs.