RFI is the level at which the ever present RF in the vicinity of an RF producing source causes a problem that is noticeable. The stray RF is always there. If a PCM receiver eliminates a noticeable glitch, that's good enough for me. I wouldn't want to fly with a predictable glitch masked by a PCM receiver. If it doesn't completely eliminate the problem with extensive range checking the problem is not really solved yet. On the other hand, an occasional minor glitch kept reliably out with PCM dosen't concern me. Of course you do all you can to isolate and eliminate the source, but it is not always possible.

Dirty Bird, I consider your RFI problem effectively solved, if you are not getting a glitch in the air. Engine-running range now double the engine-off range as you stated, is success. I don't blame you for continuing to experiment, maybe you'll get lucky. RF is very elusive and the interacting factors are probably in the thousands.

Technically, PCM radios are no more immune to RFI than FM radios. They are both FM but react to RFI differently. As I understand it a PCM receiver will not process a signal that doesn't fit within it's specifications, and waits until it gets a signal which fits the bill. Of course you have no control during this wait. If you can notice the slightest pause in control, your problem is definitely unsolved. An FM receiver will send the servos whatever it picks up without discrimination, and the results are chaos instead of delay.