I do believe that it is possible to read the signal from this ESC now. Chances are that it's a simple "1 pulse per electric pole" output, and RPM is measured by counting time between pulses.
The AP_RPM library is already capable of measuring this on both copter and plane. I had been working on this stuff in December, and currently have several helicopters flying with an internal governor mode, using the AP_RPM library, and a pulse generator on the main shaft. The governor part of the code is not in Master, but I believe my improvements to the RPM library are.
There is still a problem with the Pixhawk RPM measurement code, but it only surfaces with very low frequencies, below about 500 RPM on a single pulse/rev. I think it's a timer overflow in the pulse catch interrupt code. Something like that.
I do not know what the maximum RPM measurable is.
Note, that we can only do 1 RPM channel direct into Pixhawk.
I had written code for Teensy that can handle at least 4 channels of data, and RPM's up to I think 100,000 rpm, and it output data over I2C to the Pixhawk. So this could have been used on a quadcopter, and probably expanded to 8 inputs. Unfortunately, none of this ever made it to Master either.