The necessity to cool the sample is something I've never really understood.
If the change in pH with change in temperature is linear, why couldn't you just adjust the mash temperature reading?
Also - if the above is feasible, would you calibrate your sensor at mash temp?