It is miles per resident. So people driving across Wyoming add to the numerator but not the denominator. You can see similar effects in Oklahoma, New Mexico, Indiana, Missouri, Vermont, and Alabama where people from more numerous states drive across or to them on vacation.
But would we expect increases in vacation traffic by road between 2005 and 2011 that would have a measurable impact and push numbers higher?
Many of the states that had high numbers because of low density next to populous numbers (eg VT, ME, WY, OK, WV) had declines. The effect is not related only to vacations, but includes trips into the neighboring state as well.
LA would have a Katrina effect for 2005.
I think Alabama is due to non-comparable data. AFAICT, the data is based on state reporting to the FHA. Alabama had a big jump between 2009 and 2010, and it was concentrated in urban minor street (minor arterial, collector, and local). I think Alabama did something in their data reporting.
Incidentally, rural mileage peaked about 2002. urban mileage plateaued around 2005. From 2002-2011 rural mileage declined 14.1%, urban miles increased 5.4% with almost all that in the early part of the period.