While there are some interesting use cases in his presentation (embedded below), I am not sure that they will necessarily achieve success overnight. While it would be great to have a standardised solution for applications that rely on proximity services, the apps have already come up with their own solutions in the meantime.
The dating app Tinder, for example, finds a date near where you are. It relies on GPS and I agree that some people would say that GPS consumes more power but its already available today.
Another example is "Nearby Friends" from Facebook that allows to find your friends if they are nearby, perfect for a day when you have nothing better to do.
With an App, I can be sure that my location is being shared only for one App. With a standardised solution, all my Apps have info about location that I may not necessarily want. There are pros and cons, not sure which will win here.
Anyway, the complete presentation is embedded below:
For anyone interested in going a bit more in detail about D2D, please check this excellent article by Dr. Alastair Bryon, titled "Opportunities and threats from LTE Device-to-Device (D2D) communication"
Do let me know what you think about the use cases.