Ok, I'm quite pleased with myself about this one. It works
suspiciously well for something I've never seen done this way in any of my stepper research. "Measured torque" from manufacturer datasheets:


Inputs to torque curve generator:
- PSU voltage
- Motor rated holding torque (specify whether it's two-coils on or single-coil on)
- Motor rated current
- Motor coil resistance
- Motor coil inductance
That's it! Nothing but common datasheet values! Boom!
The basic methodology is to calculate an "adjusted" inductor time constant for the coils, taking into account supply voltage and predicted motor back-emf, and then calculate what % of the target current can be reached within the full-step period.
Of course, it won't find resonance zones or driver problems, but I'm kind of shocked how well this seems to work. Frankly, I think it shouldn't work this well, since it's a hacky technique and the motor specs are usually +/-10% to +/-20% tolerance. Really need to check more motors that have published torque/RPM curves.
I'll clean it up and upload it to StepperSim soon.