Differences between both would reveal more about commutes.
It'd be cool to see tank-by-tank mileage differences, so I can see fuel efficiency effects of traffic fluctuations (like holiday traffic weeks v. home ballgame traffic weeks).
That's a cool idea. Like "I get better gas mileage on Tuesday's when i leave for work 30 minutes later/earlier than usual. I'll try and convince my boss if I can do this all the time".