I've never tried it myself, but when talking to both able-bodied people and people with SCIs who have tried it, they don't report it being any more uncomfortable than regular FES. It can result in a much stronger muscle contraction which some might find uncomfortable, but it doesn't have to depending what your goal is. It also is quite important to consider what those stimulation parameters do to pain-specific afferent fibers, which I can't comment on as I don't know enough about them.
I'm thinking of what I've read from the RISE project, which pioneered wide pulse stimulation for denervated muscles. They only recruited subjects without sensation.
"Grading of the electrical stimulation-inducedfunctional behavior of the complete LMN denervatedhuman muscles needs a special classification since thecomplete LMN denervated subjects are and ought toremain without sensory function in the thighthroughout the study to tolerate the high currents of theFES training for complete LMN denervated humanmuscles."
Granted, they used pulse widths up to 150 milliseconds with intensities up to 250 milliamps. Though this begs the question, does stimulation with only 1 millisecond (or 3 millisecond, in RTI's case) with normal intensities (up to 140 milliamps) provide the same benefit as what was reported by the RISE project?
Good god! How did they not burn right through the skin with an intensity like that? With denervated muscle I suppose you need that kind of strength to activate the neuromuscular junction directly? Either way that's insane.
From what I've been told about restorative therapies and other devices, the amplitude does not need to be nearly that high, and it shouldn't be, if you have a fully intact reflex loop. This is a much different case than fully denervated muscle.
H-reflex testing is technically uses wide-pulse stimulation, and that's done routinely in research on able-bodied people and people with movement disorders all the time.