A measure of efficiency commonly used in the aerospace industry is "seat miles per gallon". I think to factor in weight, we need to compare "actual payload-mile/gallon".
i.e. examples:
If you drive a 5800# SUV, get 13 mpg, and normally drive by yourself at say 200#, thats 2600 payload-lb-miles/gallon. Weight of vehicle shouldn't matter, it's the weight of the payload carried that matters. On the other hand, if you filled all 9 seats with 220# passengers, and manage 12 mpg, you get 23,760 payload-lb-miles/gallon.
A motorcycle at 70 mpg with a 170# rider would be 11,900 payload-lb-miles/gallon.
A 12 passenger Dodge Sprinter Diesel full of 200# passengers at 26 mpg would net 62,400 payload-lb-miles/gallon.
just my .02
__________________