Just to agree with what others have already said, I don't think it is possible for time to run that time. Here are the best results I can find from Mike to date:
5k - 21:52
35:54 - 5mile
47:40 - 10k
1:40:44 - 1/2 marathon
1:14:48 - 10 mile
I used Merv's running calculator (not completely accurate, but still a good predictor) to see what each of these performances "is worth" at the marathon distance.
http://merv.stanford.edu/runcalc
5k - about a 3:34:15 marathon
5 mile - about a 3:31:24 marathon
10k - about a 3:42:28 marathon
half marathon - about a 3:31:29 marathon
10 mile - about a 3:29:49 marathon
So the 10 mile appears to be his best result, but none of these at all indicate 3:11 is even possible. If he ran a 3:20 I may believe it, but even that would be a huge statistical outlier. Since his Boston time was in-line with his other performances I am not inclined to believe that a miracle happened here.
Basically, if others on this board performed as much better at the marathon (relative to other distances) as Mike there would be a lot of Olympic Champions here and maybe even a few sub 2 hour marathoners.