Alan:
Although I have never seen a missed impulse, I see that you are still holding to that theory. I suggest using two Proteges because not only would allow a check for spurious impulses, but would also check for missed impulses. Of course comparing with a Jones would do the same thing.
Incidently, with the two-magnet method, a gain or loss of an odd number of impulses shows up by the fact that the meter starts to increment at a rim reading of 0.5 instead of 0.0.