Basically, one runs a current (usuallly constant, and small) across a conductor (often this is really more of a semiconductor material). There is very little voltage drop across it until a magnetic field oriented perpendicular to the current flow is applied, which causes a voltage rise across the width of the conductor, which is used as the signal to generate ignition timing in this particular application. Actually, in this switch, the magnets cause a voltage to appear across the conductive material until a magnetic/ferrous vane passes through the air gap and basically "short circuits" the magnetic flux between the 2 sides of the sensor. This is what causes the "pulse" that is fed to the ICU.
All open collector output devices need a pullup resistor of some sort - it is possible that there is one integrated into the ICU on the bike, but I don't have a schematic for it. It shouldn't be too hard to test one to see if it needs an external resistor or not. The thing that I noticed when I was looking into these awhile ago is that the sensitivity and responsiveness of the device is a bit different from the original Honeywell. I think that it would probably still work, but it might have to be tweaked a little in comparison to the Honeywell device so that the firing angle is exactly the same - I think by my initial estimates the difference would only amount to a few arc-seconds..
