I think if you'll take another look at my comments, you'll find that I did say that transient noise on the power line can produce noise in the audio. My point was that surge suppressors as commonly used cannot do anything about noise of this kind. This is what you found with your "unsuppressed power strip" which may or may not have been a surge suppressing strip.
Ferrite traps are quite commonly used with computers, but unless one is provided with a component, sized and located appropriately, debating the addition of one to an otherwise "unfiltered" component is seldom productive, hence my not suggesting that in the case in question they could be purchased at Radio Shack (if you could expect to find anyone there with sufficient knowledge to figure out how to order a fistful for you). Other than brute force "put one everywhere" methods, figuring out what size/kind/construction/location "trap" to apply is not exactly simple, and most such "amateur additions" will have little benefit or effect.
The entire description of the noise here is "the pop pop pop even when the volume is turned all the way down." That's not a whole lot to work with for purposes of designing a noise suppression scheme.
The "problem" described is that off/on cycling disables computer function(s). The correct "solution" here is that "random devices" should not be disconnected and reconnected without expecting the need to reboot when connections are changed.