It would be quite difficult to measure the actual long-term speed stability differences on a crystal-controlled device. My feeling is that this particular measurement would be rather meaningless. It would be possible to measure the W&F however that measurement too has its limits as it depends on many variables, including the test record used.
Alex

