It would increase per-driver revenue, certainly, but probably decrease Uber revenue fairly significantly. Taxi services, whether in the old fashioned way our the glitzed-up smartphone way, are pretty commodity.
The success of budget competitors like Lyft and Sidecar is testament to this. The perpetual expansion of UberX too. It seems most people are more concerned about getting from A to B cheaply than any kind of premium lap-of-luxury service.
A ride-splitting service would be pretty great for consumers and maybe for drivers, but from Uber's side it would be a pretty substantial price cut.
Great for the driver, great for the two passengers, bad for Uber.
For the $16 trip, let's say UBER gives $10 for the driver(instead of ~$8), and takes $6. UBER still comes ahead. And don't forget the huge lockup it gets - which is valuable in and off itself, esp. in the long term as a basis for self driving cars service.
There's of course the issue of decreased traffic for drivers, but there's probably a strategy to make it work.
I think this is a totally different business modal. I am not sure if many customers would like sharing a cab with a stranger.
But it happens in Beijing, especially when the demand is super high (e.g. 2 hours queue in a rail station). Well at least it happened to me :)
Such option could easily increase demand, be more cost efficient and higher margin and be highly dependent on market reach , which makes it very difficult to replicate.
And i'm sure UBER has some data to demonstrate this and other attractive options to investors[1], which make it much harder for us non-insiders to truly evaluate UBER.
[1]A similar example is the nest acquisition, which raised many eyebrows and guesses, but it appears they had a "secret" way to monetize their service - selling demand response services to utilities.