Yeah, the resonant waves tend to be surface flexing (the top side expanding the bottom contracting and thus the sheet as a whole bending; and ofc vise-versa) which has a natural frequency related to the lever arm length amplifying the moving mass vs. the more centrally located elastic crystal lattice deformation forces acting as the spring.

At least that's my recollection on the mechanics of these types of resonators; the general frequency of period being proportional to length and thus your observation holding is fairly universal, though.

The 32kHz tuning forks btw use a torsional vibration mode thus not needing to be as massive as linear scaling would leave you to assume.