Oh that's interesting and, in my head, it makes sense the drive power would cause the frequency to shift, I think, more drive, larger excursions, lower frequency because of greater movement.
Am I being simplistic or just plain wrong?
I don't think that for my application the lower sensitivity will matter too much but I'm planning to build at least two devices so I might swap the 'microphone' a few times and see if I can measure a difference.
I've just ordered some MEMS microphones and am playing with Teensy FFT code, I may end up with a BATSDR...