For these reasons, I think that GM will be the big fish when the dust settles.

GM has indeed the best position to gain a large market. The problem with this technology is that it is moving at blazing speed. I've tried to limit my predictions as much as possible for that reason. GM's goal is to get a Cadillac on the road with this technology in 2017. Assuming that works, it will be at least a year, possibly two until most of the GM fleet has the same kit of sensors, and they will definitely start at the higher end because this tech ain't cheap just yet. So we're talking about 2018 or 2019 for mass-scale production at the very least.

Just two years ago, Google's cars couldn't handle roundabouts, railroad crossings, unpaved roads and most busy areas without failing. Tesla hadn't even started working on their Autopilot features yet. I don't think I can make good predictions for three, four or five years in the future. Especially since Tesla is working on the Tesla 3, a $30,000 car for the masses. They might just beat GM to the punch.