Lynx

Actually, twisted wire is longer than the same measured length of straight wire. Now in theory this means your twisted wires cost .001 cent more per wire and your battery time is reduced by .0001 seconds per flight because of the increased resistance <G> But that's only if you wanna be zany about it (Note to serious geeks out there, those numbers were NOT calculated so don't try to prove em wrong)

The signal is AC, but it's at... (you guessed it) the servo refresh rate of 50hz, so you're not going to prevent any RF interference FROM a servo by twisting your wires. And unless you run your antenna perfectly parallel to a servo wire there's not likley to be any measurable crosstalk. Though with the pulsing current of a servo near the antenna, common mode DC shifts in the receivers signal might cause problems with range or general reception depending on the exact electrical setup.

Twisted wires are usefull, that's why UTP cable is so common, but those are only usefully realized if you have a differential signal, and it only usefully cancles out RFI at higher frequencies (which RC'ers do not use)

The problem is if a wire is carrying current, any wire next to it will try to develop a voltage in responce to that current because of the magnet field the orginal creates. Now if you have a servo near an antenna wire they draw a LOT of current (comparativly) something in the neighboorhood of .2 amps to 1amp or more. And when you have a servo under a lot of load in flying conditions (being buffeted or dramatic stick changes) The current going through that wire is varying extremely rapidly. This screws things up if they happen at a frequency that the receiver isn't capaable of dealing with.