Looks like the Eagletree would do the job. Though the price + sensor is pretty much the same as a new receiver.
The only real reason why I was interested in RPM was purely for gearing. i.e. trying to maximise the duration the motor is at peak torque, power and efficiency (maximising 'area under the curve')... but mainly tuning for peak power.
I'm not really keen on only tuning by temperature. Each motor has different thermal characteristics and is affected on ambient temperature, humidity etc. i.e. too many variables. Tuning by rpms seemed like a more direct and consistent method.
It seems the only other option would be to glue 2 magnets to my spur gear and use a hall effect sensor with some kind of custom made circuit or arduino to find the peak speed - then do some basic math. Seems like a lot of trouble for a simple problem - It seems the only other option is to go by ear or "feel" as they say, which is not scientific at all :/