Looking at the MX500 schematic, I'd tend to think the appropriate way to fool it would either be to add a 100uH inductor between out and ground or to short out C36 (possibly with a SPST switch -- although you'd want some way to access it (i.e. drill a hole in the case -- not sure that would be fun). A switch shorting between U2.1 and U2.8 might be another interesting option as that would not only disable the check but it would reset the latch -- which would mean you could change tips without shutting it off to reset it.
I don't see a way you could simply add a resistor since you don't want it in series (and it wouldn't even do any good because there is DC blocking cap in the wand), you could shunt it to ground, but unless its in the DC path (in which case, just short out C36), it'll shunt all the RF through it.
I'd still rather modify the wand just in terms of a "cleanliness" thing, I don't like the idea of losing functionality -- but maybe it doesn't matter.
As for the collet thing... GreyWoolfe, perhaps you've never used the RM2. the female barrel threads on just like any other RF connector. I've never had it pull off, it goes on easily, and tightly. The RM3E connector (and the DS1 tool, etc) push on, then you tighten the collet, but even then its not super tight, so you twist and twist and twist (and have to twist the whole cable) and it never gets as tight as the RM2 connector. It seems the collet doesn't tighten enough. If it did fully tighten, then I'd agree it might be a bit easier to use (sort of the difference between a quarter turn valve and a multi-turn), but the collet doesn't seem to tighten adequately. Just to let you know, I have an STSS-001, MX500, RM2, DS1, and I borrowed a friend's MX500 and RM3E. All the collet based connectors performed similarly on both MX500s, so I assume its a design based issue not my particular unit.