Can't say for sure, but I have been able to drive a servo on a breadboard test basis with no sign of jitter, so it can be made to work.
I would think the most likely candidate is power or ground noise. A scope would be the best tool to try and see what is happening electrically to either the power, ground or signal lines on the servo.
If you can temporary hook up 4 fresh AA alkaline batteries in series for the servo power that you give you a quick way to eliminate your external power supply as a problem or not.
PS: I would not attempt to use opto-isolators for the servo control signal, not need and might effect the shape of the signal pulse somewhat.