This approach to rate analysis is interesting. I know from experience, having seen how hard it is to deduce correct truckload rates for a given simulation. We spent a lot of time divining these rates for a specific need. A generic econometric approach would be useful. I hope Chris Caplice has a public access paper on the subject.
The same issue applies to rail rates as well. And they are even more obscure, since one cannot get info on them easily.