Re: And next: commercial positioning
I'm not sure that Starlink _is_ compatible with a GNSS application. I think there are three main issues:
-- As I understand it, the individual Starlink spacecraft have some degree of autonomy in order to avoid each other. So a given article may not actually know, precisely, where it is! Maybe it will be good enough, but...
-- I also doubt they have a highly accurate clock, because that adds cost, and as the whole point of the constellation is to maintain low-latency connections, whatever timing information a given spacecraft might need would be amply addressed by an IEEE1588 clock synchronization system.
-- The terrestrial RF links are narrow, with the system working like a cell system handing off "calls" to different access points/satellites and beam-forming used to solve the power problem. This means that the footprint of a beam may be tiny, e.g. a 10km circle.
So a classic GNSS is probably not on the cards. But if a Stalink satellite "knows" enough about a ground-station's location to steer a beam to it, that would presumably make it theoretically possible to get the position of that 10km circle. And successive spacecraft could, over time, improve on that ... but this is all assuming that the system doesn't already depend on GPS to steer the beams to the ground stations (which may or may not be the case, but I wouldn't rule it out!).
What I do think is very plausible is that, if there's a good business case, that there may be a GNSS variant of the Starlink spacecraft. One very specific business case that I know SpaceX has put thought into is... how do you locate yourself on Mars? If the answer to that is a Martian GNSS, an Earth variant is an incremental and useful addition, and Starlink proves they have the technology to launch and manage such a constellation...