I've found dynamos for our engines to be rather finicky.
With my diy electrical testing rig, I have 3 DC motors (generators), a bulb and a Wilesco LED.
Using my Gilbert motor gearbox as a steam engine stand-in. From experience, the Gilbert engine isn't as powerful as a Wilesco D16.
The most reliable wattage output came to 0.22 watts from a toy motor. Interestingly, the Microcosm FD4 was all voltage and not much amps.
They are so sensitive to the right gearing... Or windings, field strength and rpm. I wonder how the PM Research dynamo would do?
Interestingly also that the load makes such a difference as well.