The design will not take off due to the limits of consumer routers (as well as many enterprise ones. Wifi throughput drops significantly when you add more clients, even if they are all running at the same mode with the same QAM. Other than that, if you happen to end up with a mixture of clients, then performance takes a massive drop.
On top of that even if you are not using wireless for your internal network, the act of allowing traffic from other people, increases the workload on the router, and benchmarking applications can detect the impact on other traffic. even if you are not using the full throughput that the router has to offer, the more network traffic you have, the worst the response time gets (this is noticed on even high end consumer routers such as the R7000 and the WRT1900AC. You generally do not escape this for a large portion of activity until you get service provider level equipment, or other high end enterprise level networking equipment.
overall this will not be effective unless they can make a router that can rival the performance consistency of high end enterprise level networking equipment, in addition to having a dedicated wifi radio just for the public network. They must then develop a new form of QOS that will allow overhead free bandwidth management (current QOS has some performance hit, in addition to negative impacts on the response time of traffic going through it.
On top of that, wifi is still largely a collision domain such as with the classic ethernet hubs, so there are many things in the way, preventing the public hotspot from being successful at a residential level.