On the presumed coevolution of brain size