Tokyo highway30 MPH traffic would require a prohibitive number of small cells for carriers to do it themselves. Wi-Fi First is not a complete substitute for LTE towers. There will always be spots not covered because small cells, especially in higher frequencies, have very short range. Below, the CEO of American Tower puts forth his opinion why highways and moving vehicles are particularly difficult to cover. James Taiclet via Seeking Alpha.

Taicet. The handoff requirement from places where our towers serve people, which are often around highways and other transportation corridors, suburban or rural, you've got people traveling 30 miles an hour to 60 miles an hour. You can't really have sufficient handoff capability over a very large stretch of multi-mile roadway to economically provide those handoffs. 

Taicet, thinking of carrier small cells, went on to say the economics become dubious with densities of less than 10,000 people per square mile. Wi-Fi First changes that, of course.

You've got to have a fiber connection to every small cell, so if you're going to try to cover the roadway from Hopkinton, Massachusetts, where the marathon starts, all the way to Boylston Street, you need hundreds and hundreds of small cells to do that. You'd need 26 miles of fiber just to do one road. And that's one of many, many roads that go from west to east in our area. It just is an economically infeasible opportunity. And you also need, by the say, siting costs. Wherever you put your small cell, you usually have to pay somebody; whether it is the town, the utility, have a revenue share.

These problems are solved by using the customers' home gateway, with 14 million hotspots already connected in Europe and Japan. Backhaul's built-in. No cost for location; most people are happy just to get coverage in turn when they are out. (Must be strictly optional, of course.

Wi-Fi First can be revolutionary.