Resistance to interference is "all relative." Even the best fully professional equipment can never, on its own, be absolutely, 100% resistant to all interference. People need to realize that it's just asking too much of the equipment.
In general, the setups that are safest from interference are those that are balanced from end to end. All other things being equal, balance is more important (effective) than shielding. But balanced setups can have "pin 1" problems, depending on the design of the equipment you're using (which is why Neutrik "EMC" series XLR connectors were invented and should be used much more widely in my opinion). And ultimately, even balanced setups with no pin 1 problem can still suffer from interference if the source is strong enough and nasty enough in terms of its spectrum and spikiness.
With an unbalanced setup you are just a sitting duck, and the only defense you have is your hope that no offending signal sources will come within range.
The distance formula (inverse square law), cited above, sets the stage. But the mechanism of interference is that the RF signals are rectified (detected) across some junction in the circuit and then amplified--and detection is a threshold phenomenon in terms of voltage. Given any particular source of interference and any point in a circuit where it might be detected, if you're beyond the corresponding threshold distance, you're basically safe as long as the signal stays the same. But if the interference source gets nastier and/or moves closer to the point of detection, the resulting interference will increase disproportionately as the threshold voltage is crossed.
So you really do have to keep cell phones and other mobile electronic devices away from the equipment and the cables--all the more so when unbalanced equipment is being used. Shielding is nice, but niceness doesn't help much against a signal that's basically like a rapidly spinning Ninja blade being thrown at it.
--best regards
P.S.: I don't see how the storage medium (SD card, etc.) could be an issue, since by the time any writing occurs there, the signals have already been digitized. Even if the interference were somehow strong enough to disrupt the data writing process, the result wouldn't be an analog of the detected interfering signal.