Right now it's in the experimental stage, with only 6 towers total deployed (only 5 were operational during NAB, and only one in Nevada... so timing, not navigation yet).
The ultimate plan—which is probably dependent on how well ATSC 3.0 rolls out (which has plenty of hurdles[1])—is to encourage broadcasters to add on the necessary timing equipment to their transmitter sites, to build a mesh network for timing.
That would allow the system to be 100% independent of GPS (time transfer could be done via dark fiber and/or ground-satellite-ground directly to some 'master' sites).
The advantages for BPS are coverage (somewhat) inside buildings, the ability to have line of sight nearly everywhere in populated areas, and resilience to jamming you can't get with GPS (a 100 kW transmitter signal 10 miles away is a lot harder to defeat than a weak GPS signal hundreds of miles away in the sky).
The demo on the show floor was also using eLoran to distribute time from a site in Nevada to the transmitter facility on Black Mountain outside Vegas, showing a way to be fully GPS-independent (though the current eLoran timing was sourced from GPS).
[1] ATSC 3.0, as it is being rolled out in the US, doesn't even add on 4K (just 1080p HDR), and tacks on 'features' like 'show replay' (where you tap a button and an app can stream a show you're watching on OTA TV through the Internet... amazing! /s), DRM (at stations' discretion, ugh), and 'personalized ad injection' (no doubt requiring you to connect your TV to the Internet so advertisers can get your precise location too...). Because ATSC 3.0 requires new hardware, consumers have to be motivated to buy new TVs or converter boxes—I don't see anything that motivates me to do so. I feel like it may be a lot like the (forever ongoing) HD Radio rollout.