I observed the same sort of relationship between temperature and pressure in my five year study of a pneumatic tire, reported in Measurement News, Number 127, September 2004.
In spite of seeing how well one can predict tire size knowing temperature, I'm reluctant to abandon post-measurement calibration. Unless one rides all the time with a pressure gauge attached, which is not easy, one does not know how the tire is changing. Does it have a slow leak? A postcal will tell you.
As for using remote calibration courses, if the calibration course is so far from the race course that it is inconvenient to return for a final adjustment, one should simply lay out a 300 meter or 1000 foot calibration course on site. This is always a good idea when a measurement is away from home.
In 1999 a group of measurers descended on Pittsburgh to do a group ride of the US 2000 Olympic Trials marathon course. I was the data taker, and did not ride. Results were reported in Measurement News, November 1999, Issue #98.
Part of the data I took was tire pressures at calibration time. In order to not disturb the tires’ calibration, they were initially pumped up to a desired level. Then I used a digital pressure gauge (Smart Gauge by Topeak) to read the pressure. Once this was recorded the bicycle was calibrated. The tire pressures were again checked after postcalibration was done.
One thing I noticed was that if I was not extremely careful, the seal between gauge and tire valve did not always get made or broken instantly. A short hiss indicated that a shot of air had escaped. When this happened the pressure was reduced, and the original pressure was forever unknown. Although I tried hard, I had enough bad tries that I judged that my pressure data was too flawed to be meaningful.
Using a pressure gauge on a tire always causes a small amount of air to escape from the tire to fill the void within the gauge. Thus any gauge reading will be lower than the actual pressure before the gauge was applied, depending on the amount of air that is allowed to escape.
I think making adjustments based on tire pressures is unjustified simply to avoid recalibration.