Pretty obvious what the move is.
With this program, Google is trying to find a way to build up a critical mass of cellphone users who paying Google to be the "middle man" between them and their cellular data provider. Then, in turn, Google can approach any given provider and buy network access in very very large chunks.
Once that critical mass is reached, or even while it's being reached, Google can correlate the usage data with location data (gathered from your phone) and replace all the high-volume cell traffic regions with their own fiber wifi hotspots instead.
So, slowly, they leech data usage away from the other carriers, while at the same time building their own network, starting in the most important areas and working to the fringes.
Eventually, Google will declare complete (or almost complete) independence from the carriers, and - ta-da! - turn their inferior cellular service into a highly competitive one, with the advantage that now, they can data mine you as a man-in-the-middle, simply by sniffing the traffic on "their" network.
It's clever, ambitious, and right out of the Microsoft "embrace extend extinguish" playbook.
There are two problems for Google here:
1. AT&T and Verizon know what they're up to, and don't want it to happen
2. AT&T and Verizon already offer good coverage - including WIFI spots - for the vast majority of cellular users.
So Google needs to really push the idea that their service is "better" because it hops between Sprint and T-Mobile, hoping their users don't realize that Sprint and T-Mobile are, in turn, piggybacking AT&T and Verison's infrastructure. There really isn't much gain from that.