You can greatly simplify the project by using samarium cobalt permanent magnets in a rotor that interacts with a stationary aluminium disk. The amount of drag may be controlled by changing the distance of the magnets to the disk. If you need to actually calibrate it to an absolute value then an electric motor of known hp can be used to drive the dyno. Really though, the focus will be on relative engine output rather than actual calibrated output. As long as the unit can show changes in dyno input accurately that is all you need to know.
The power can be measured by reading the temperature rise of the aluminium disk or you can directly measure the torque produced and obtain an actual torque measurement that together with rpm will give actual horsepower. Measuring torque is as easy as restraining the rotation of the aluminium disk with a spring scale such as may be used for weighing fish.
Samarium cobalt magnets must be used because of the heat. Neo magnets won't withstand high temperatures.
This isn't just theory. I built a magnetic transmission for my 3 wheel cart using an aluminum disk based eddy current torque converter. It worked fine except for the amount of heat generated. I didn't have samarium cobalt magnets so the efficiency fell quickly as the magnets heated up.