I think I understand what you are after, - some measurable effect from the deformation of a cavity due to acceleration. Because the cavity will be loaded by your various circuits I am not sure that it will be very sensitive. two things come to mind, the first is to use the "flank" of the resonant circuit, which may fall at 20db/octave, so a frequency change of 1% would lead to a level change of 1%. The other is to have a second cavity and you cancel the two outputs of the cavities at the resting frequency hopefully as one cavity deforms, the balance will go and a large output will result.
Frank