I wonder if you make a 3 axis MEMS device, and then use it to generate the compensation coefficients for the TXCO if that would allow the oscillator to be more accurate, down to the 0.01PPB level, as opposed to the current Tek one of being compensated to 0.1 PPB, which is an order of magnitude better than the original unit.
Then you can compensate for it being used anywhere in the world, and it might even be possible to compensate it for use in space, as there you do need to compensate for Einsteinian effects on clocks, as they do run slower in orbit. If they need somebody to hold it in the ISS I am available, just not tomorrow as I am busy all day, full day already set out and then I might even be able to do my own work a little.