Let me start off by digressing.
There are notable cases where electronic components are vibration tested. Satellites are designed and built to survive two launches into space: the first being a ground-based test in a high-intensity noise testing facility (see
http://iar-ira.nrc-cnrc.gc.ca/smpl_9c.html); the second being the launch itself. In that case you are not really testing how the device performs during the period of high noise / vibration, but rather whether it shakes itself to bits or not on its way into space. (I've seen a picture of a standard Sony videocam that was accidentally left in such a test chamber during a test. It managed to shake itself literally into all of its constituent bits and pieces to the point where it was not recognizable as a camcorder.)
Another bit of anecdotal observation: I have two lamps that I sit on two of my subs. I tend to have to replace the bulbs in those lamps much more frequently that in lamps that are not being vibrated. However while they are operating, there is no change in the actually performance of the bulb.
So my take on the topic is that vibration with respect to solid state home theatre components is to be avoided where the level / frequency of vibration may caused physical failure due to fatigue effects. But I doubt very much that a vibrating component will sound any different than one that is not - transducers (like turntable pickups) and components that "rattle" excepted.
Jeff Mackwood