RANDOM TREES AND THE COMPARATIVE METHOD: A CAUTIONARY TALE