This does not really seem to be about measuring the earth; it's about more accurate GNSS. Then if you have very accurate GNSS you can also get more accuracy in the ITRF.

Also to add context: there are many hurdles to hitting a 1mm accuracy for a satellite based positioning service.

One can already achieve accuracy of a few cm, but it's a painful process. You basically place a receiver somewhere, keep it there for a few hours, then download the pseudoranges, and post-process on a computer. In post-processing you correct for ionosphere error and satellite ephemerides (the satellites broadcast their rough location and clock readings, but you can download much more accurate data for these later to get a much better idea of where the satellites were). You need to do post-processing after some delay, because ionosphere readings and satellite ephemerides are not available immediately (the accuracy increases as you wait longer!).

You can also extend this with what's called "Precise Point Positioning" (PPP) using one ground-station that broadcasts an extra signal to moving receivers nearby, so there are ways to get higher accuracy on moving receivers, but it's still painful.

So probably 1mm will come soon to fixed ITRF monitoring stations (basically the ITRF tries to track tectonic movements, etc as those are at the mm level), but it's very far from coming to drones or cars.