my experiences are very different to the stated outcome of the analysis that "It does not put new drivers on the road."
I use Uber fairly regularly in London with friends (more than monthly compared with perhaps annually for "Black Cabs" prior to Uber) and I often chat with the drivers about their experiences, financial impact of uber etc. because my father is a cab driver in a city which does not yet have Uber and I have a vested interest.
When I ask about the surge pricing, every driver without exception has told me that he works around "chucking out time" (common bar closing time in the UK) more than he normally would because of surge pricing.
The methodology the author (above) uses seems extremely suspect, and I cannot agree with this conclusion that it doesn't put new drivers on the road. He seems to think that as soon as surge pricing activates, that magical drivers will appear from nowhere within 2-3 minutes - and if that doesn't happen he claims "it's not working". He is missing a MASSIVE factor here (that also affects taxi cab drivers other than uber) which is that demand is predictable - busy late-night areas, concert and sporting venues at the end of events etc. are known to be busy well ahead of time. The number of drivers in that area at that time would not be so volatile as to immediately increase when surge pricing kicks in, but it is highly likely that the number of drivers in the area *before Surge Pricing even kicks in* is way way higher than it would otherwise have been.
For example: A good taxi cab driver will choose to be in predictably busy areas a while before they get busy...
In conclusion, author sucks. Source: family in the industry, and analytical/logical brain.