I don't know if I believe this. The clear implication is that if we _really_ wanted to reduce our energy intensity, we'd all be riding motorcycles.
ETA: I've concluded that I absolutely do not believe this. I don't see how they can possibly _have_ btu per passenger mile for cars; the data does not exist, and in fact I question whether there can exist any reasonable approximation to the data. I can think of a variety of assumptions that could swing it by a factor of 3-5 depending on what you assumed. The data _does_ exist for airplanes after one date (when they started keeping passenger lists) and only much later for trains (late 1990s, IIRC). And that's assuming they are using that data, which is pretty damn questionable.