That's really interesting! There are so many ways to slice and dice the numbers. And this way has its merits too. I like it.
I guess the advantage to the 90-day approach is it compares cars in the same season (more or less). The 2500 mile approach would skew differently for people who drive a lot vs. people who don't, due to the effect of different ambient temperatures. EG since I don't drive a whole lot, I would have a "summer effect" with the 2500 approach that would last longer into the winter (keeping my avg high as others' start to drop), and vice versa.
What if it were a 2500 miles or 90 days, whichever comes first?
__________________