This quick test continues the work described here and here.
The setup is like this:
jump onto munter
Instead of jumps onto a knot, I use my hand as the brake for a Munter on an attache carabiner. Ten feet of double-strand 8mm sterling nylon rope (semi-static) is above the Munter, and I take "falls" of 24", 24" and 32" onto the hitch, by the method described in the links above.

The results, compared to the previous jumps onto a knot (click small image to see larger image):
munter_jump
The 1st jump gives remarkably similar results for the jump on the knot and Munter, but I doubt that is more than coincidence, since the knot was still "unhardened" and absorbed a lot of energy. I was not wearing gloves. I held the brake hand pretty tightly in the "up" position for the Munter. For the second jump, I let my the hitch slip about 6" through my brake hand, and was rewarded with burns on my hands. I  put on rubber-faced gloves for the third and highest jump, and the Munter still slipped a bit through my brake hand.  Overall it appears a fall on a munter -- and probably any rappel device -- will be softer than a fall on a hardened knot.