A lot of these "rolling resistance" arguments get wrapped around the axle, because some people are using the term specifically (and correctly) to talk about resistance that comes from tire deflection, and others use it more generally to refer to energy lost when a rider is on the bike. I think the latter case is more important to us bike riders, but we need a better term for that whole-system approach. The 42mm tires on my FG commuter bike (pumped to a mere 35/40psi) may well give up more energy on a smooth steel drum test, but I'm able to take rougher parts of the commute at faster speeds and arrive feeling pretty good. Increasing the tire pressure on this or other bikes just means that I get beat up along the way and end up feeling more worn out. Paradoxically, since the tires ought to be rolling faster.