Just theorizing (too much time). We are both looking at this from the electronics point of view. If you were running a low voltage, high current system, with different long wire lengths and diameters the resistance of the wires may vary? Hence each ESC will detect a different voltage, the RPM may differ? but that would be impossible to detect. I think I'm feeling very subtle changes of frequency more to do with the vibrations coming down the control line wires to my hand?
The next two paragraphs are not my work, they are copy and pasted from another forum.
"If there is a significant distance between the battery and the motor, it is MUCH better for the controller to be as close as possible to the battery, with the three motor phase wires being the longer section. In an effort to reduce voltage ripple and voltage spikes that result from being forced to have longer battery wires, it is sometimes helpful to add capacitors to the red/black pos/neg right next to the controller.
This is because the motor can absorb quite a bit of voltage ripple and voltage spikes, but the controller is much more sensitive. The key metric is to use Capacitors with a low-ESR. A physically larger capacitor with high capacitance doesn't hurt, but it also doesn't help either. A capacitor has two leads, positive and negative. Simply connect the two leads to the two battery cables, in a place near the controller."
Back to me. If there was a dry joint somewhere? that would cause major issues.