Thanks Cody.
The measured drift was well within the max spec of the part. I actually desired a lot of drift and so designed with a resonator rather than a crystal.
What lead to this testing was the need to validate my interference countermeaure plan for the product. Our receiver used a highly discriminating downconverter and we only needed to shift one units operating frequency 10ppm or so to evade interference from another unit. So I placed a couple of 1K resistors physically next to the resonator and the microcontroller would heat the resistors on detecting interference, which heated the neaby resonator, causing the operating frequency to shift and the interference to go away.
