Except that article only considers one mechanism for rolling resistance. There is also energy loss from a tire bouncing on rough surfaces (as all real roads are). Above a certain pressure (usually around 120 psi) this grow faster than energy loss in the the rubber decreases. As a result further increases in pressure cause rolling resistance to rise.

And this has nothing to do with efficiency since efficiency makes no distinction for where energy is dissipated - heating air, heating rubber, or working against gravity.