I'm suspicious of the 5K Ohm limit on plugs. EVERYTHING I see I can trace to Tempest, who uses a carbon resistor. I believe those plugs, like Champion, who use the spring and slug, are using ceramic (silicone carbide based) resistors to stand up to the vibration in that style. I seem to remember from EE school that silicone carbide is a semiconductor and that it's resistance is not constant over the voltage range of sparkplugs...but that was a long time ago.
In any case, measuring a spark plug with an ohmmeter is like doing the same to the high tension leads. They should at least be measured with an insulation tester using about 2500 volts, not a VOM using 3 volts.
Mechanically, a cast in place carbon resistor I would expect to read more consistantly with a VOM. The contact surface is large between the resistor and the rest of the plug.
Clearly a very high resistance will be a problem. But if I assume the 'distributor' and plug gaps combined (0.025"?) need about 10kV to fire in the running engine, and that a magneto can reliably put out at least 20kV, then there is 10kV available to push the required current thru the resistor. I read somewhere recently that the most any plug would need in current to properly fire the cylinder is 0.1Amps. 10kV will push 0.1 amps thru a 100kOhm resistor (including the resistance of the various contacts in the plug). If the plugs are fouled at all, the fouling will look like a big resistance, and I don't know how big.
I am interested because as I look at switching from the old C10S resistor-less plugs to the 707SR-10s I just bought, I find that the first three 707s I opened measure about 13kOhms with my Ohmmeter. Has anyone using 707s measured them???