I remember that Mike Sandford posted that he believes the change in calibration constant due to temperature change has more to do with the rubber material properties changing than with the pressure change. It would be interesting to do an experiment where you released whatever pressure increase you got due to temperature rise, and see if the cal constant still goes down.
10 years of measuring for me without having a flat during a measurement. But I suppose it only takes that happening once at an inopportune time to make you consider airless. I use a smooth Kevlar tire on my mountain bike that holds 100psi. My cal constant doesn't change much with temperature, probably due to both the high pressure and the thin rubber in the tire.