> Are you an electrical engineer, or similarly educated in RF communications, and thus able to determine what is arbitrary?
CpE here. The vast majority of communications links on planes use twisted-pair differential signaling [1], which is extremely resistant to RF interference.
I've also worked on test procedures for avionics equipment, and assure you that RF interference testing is a part of all test plans for such equipment. The types of interference we tested against were far more aggressive than those that any consumer electronics can emit.
Keep in mind that the FCC also aggressively regulates RF emissions from all consumer electronics devices. Testing against FCC regulations is an involved and expensive affair, and the vast majority of devices are forbidden from any powerful RF emissions.
> if I am unsure a device that emits RF will affect my instruments, I shut it off. Inconvenience does not outweigh safety, even / especially if the effect on safety is unknown. The worst case scenario, although likely minor, is simple not worth the risk.
If phones or other complex electronic devices truly can interfere with avionics, their presence should be forbidden in the passenger cabin. It wouldn't be hard to detect them during security screening.
I conjecture the reason that this hasn't happened is because the risk of such interference is a practical impossibility for the reasons stated above.
1. http://en.wikipedia.org/wiki/Differential_signaling