It also depends on ho the Constant Rotation servos are wired.. Even though I saw a plan for converting standard servos to C-R type, using 2x 2.2K resistors wired in series, across the old POT pins, with the center bridge to the wiper solder point, I never liked the idea of sloppy tolerance on the resistors. (and, isn't 2.2K a bit low? I thought the POTs were closer to 10K end to end?)
I've done this with modified HiTec & Futaba servos, grinding down the stop tab, and replacing the POT with a mini PC mount 10K, accessible through a side hole... Even replaced the two C-R servos on an old BOE-Bot that the motors had begun screaming.
The servo code works with the VEX Motor modules (essentially Futababa servos with the post ground down, and a square key in the center, but with a custom chip for the driver).
Unlike the VEX controller code to motors, there is no constant pulse going to the signal pins, just for the brief moment you digitalwrite to the pin, or servowrite to the pin.. Once you finish that instruction, the servo would stop anyways, because you are no-longer sending the pulses to it. just a dead GND signal. Same goes with the Basic Stamp, once you write the pulsout, there is no constant signal.. the delay, just gives the code time to stop, while the servo performs the movement, then waits for the next pulse. (with the VEX, if you wrote out to the motors, the 2nd PIC inside would maintain the pulses, until you told it to stop, which lead to a lot of run-away problems if the code crashed.)