ORIGINAL: tkg
The resistor acts some what like a resistor plug. I'm not really sure of why things work as they do, but doing range check you can see what the result is of adding resistance and/or shielding.
The ancient standby has always been a 1w 10K resistor in the lead. Still works good today.
There are two parts of noise one is the spark itself, a spark has an infinate number of harmonics. The other is these ignitions don't have a smooth steady current draw. They take a big gulp of power just after the plug fires to recharge the HV cap. This puts a noise equal to RPM on the battery leads, switch and the battery its self.
Modern raidos are so good that we can get away with many sins in our installlations
I have always wondered why a resistor in the spark lead reduces the RF interferance. I think its because it limits the rise time of the pulse that creates the spark. That would reduce the frequency of the generated RF below the bandwidth of the receiver. Have you looked at the frequency spectrum of the spark with a spectrum analyser? With and without the resistor?
Recharging the HV cap should generate RF well below the frequency bandwidth of the receiver. It would be a good idea to check this with a spectrum analyser as well.