Found the bug in the code and started exercising the stepper on the plasma cutter Y axis.

Kept winding it faster until it clacked out in one direction, forcing me to back off to a mere 7000 steps per second. OTOH it did manage the return stoke at 7500 steps/second.

7000 per second is 17.5 revs per second geared 2:1 down on to 12 tooth, 1 Mod pinion.

Sounds like 13"/second to me. 65 ft/minute.
AND... only one pair of coils active in the stepper