Fair point - I think the monthly state data are less reliable than the national annual data which incorporates the moving surveys as well as fixed data points.  However, as with oil data, we do the best we can with what we have...
Oh, I'm sure we can draw some general information from the data ... but I just have a bad feeling that the data, and the relation to GDP, are too self-referential.
I don't see any reason why there's any contamination with GDP stats.  You can argue there's sampling errors in the VMT, but the correlation with GDP is evidence that there is some useful information there, rather than the contrary.