On the limits to the long-period method in classical economics. A note