A group of simulated driverless cars negotiating a typical urban intersection at the same (slow) acceleration of a commuter train increases the time needed to cross the intersection by anywhere from 36 percent to more than 2,000 percent.
If anything this will always be a strong argument against cars, even if autonomous driving takes off. Still would like a PRT for dense urban cores that don't have 24/7 transit when less than clockface intervals don't make financial sense or hard to navigate terrain like WVU.
