25 years into this RTO experiment we still have no ability to say dispositively that it has either worked or failed with respect to costs. Someone – maybe a federal regulator – should do an independent study, the author writes.