RNAV is here to stay
I would be more concerned with the stuff supposedly attached to the ground for calibration and serviceability, ILS, LOC or VOR.
The beauty with RNAV is that it does not rely on NAV aids attached to the terra not so firma. So opposite to your conjecture it would actually improve the case for RNAV, as the RNAV approaches are unaffected by ground equipment.
Land movement under the approach is the same issue no matter what sensor is used for an approach.
Normal RNP AR approaches would not be affected as the DA is such that it is slightly higher than CAT 1 and a few meters shift is insignificant (it still does better than VOR!) Differential approaches are a different animal and I would suspect have similar calibration and survey issues to CAT 2 and 3 ILS.