Guessing, but it is possible that when those ESCs were designed and developed, available hardware determined the operating frequency. Offering a switching frequency in the wrong band to a motor with a high pole count could result in the on pulses being offered out of sync with the position of the brushes and commutator, resulting in stuttering.
Since high pole count motors were not common in the model supply world, not a serious problem. Unless the enterprising modeller decides to use motors from car scrap yards intended for heater blowers, radiator fans, electric windows etc.
When messing with my home made (train) speed controls back in the early '70's I was constrained by the real world and what I had available. Going for a low frequency, the parts to generate that low pulse rate were unfeasibly large. Going for higher frequencies, I didn't have easy access to test gear that would allow easy testing. Like much engineering, the result was a compromise. The resulting constant 1KHz whistle was sub-optimal, but not too bad in a model diesel loco.
The same problems would have faced any designer 20+ years ago, but rather than the physical size of components, the limit would be processing speed.