Matt, the ESC is indeed a DC device. There is no AC current in it. I have designed 3-phase brushless motor controls too so I understand quite well how they work. You are correct that the ESC turns the switches either on or off, not in-between. Because of the dynamics of the motor being attached though, the current doesn't actually start and stop instantly when the switch (mosfets) turns on or off. This is why the ESC can smoothly regulate the speed when it is in part throttle mode. The ESC is 'chopping' the voltage rapidly on and off at part throttle but the motor reactance smoothes it out somewhat so if you monitor the current waveform the slopes are ramped and the corners are rounded. Also the mosfets have parasitic diodes within them that permit some 'freewheeling' of the current even when the switch is shut off.

My comment about the innacuracy of your prior post was in regard to your response to Bill Mixon's statement when he said he thought the motor runs on AC current. You said "Turns out that even though the ESC does switch the polarity back and forth on the various phases of the motor to make it turn, it's still putting out DC voltage. " This is wrong and self-contradictory, because it implies that the output voltage polarity alternates and yet is constant and consequently that the current will be Direct Current and constant also. I'm not trying to pick on you but rather just trying to provide an accurate explanation for everyone's benefit.

The motor DOES run on AC current. In fact it would be very happy to run on a 3-phase AC current with a sinusoidal waveform like that of industrial 3-phase AC (household AC is single-phase), so long as the phases were synchronized to the rotor position. Some high-end brushless motor controllers used in industrial applications actually do synthesize a virtually sinusoidal drive waveform using DC electronics essentially the same as our ESCs, by modulating the timing of the voltage chopping. The sinusoidal current drive is the most efficient way to drive a brushless motor but it is more complicated to design the controller so they don't bother for our ESCs. It could be done in the ESC software though so don't be surprised if you start seeing sine-modulated ESCs soon.

Unlike a brushed type DC motor, if you were to spin the shaft of an unconnected brushless motor and display the output voltage waveforms on an oscilloscope, you would observe nearly perfect AC sine waves. You could use it as an alternator. To get DC out you'd have to rectify the current with diodes.

It is a bit confusing to try to understand that even though the ESC is DC and trying to chop the current into nice square waves, the actual motor current that results becomes not-so-square AC.