Stu,
Probably the best place to go to suggest this idea is
http://www.diyaudio.comI'd suggest that you post it there - you are very very unlikely to get any frivolous replies there, and there are some very smart and experienced people there...
Magnorestricition is known, but the question remains as to how this mechanism in a given cable of a geometry not prone to produce electrical variation would operate?
Wire "effects" are reported in cable geometries where in the example of speaker cables the two conductors are not physically associated with each other by mechanical connecton... so are you proposing that wires that are in parallel running the same signal would produce a magnorestrictive effect WRT each other, and that this would in turn develop signal in the conductors?
And, in addition it is generally accepted that the place where "differences" are heard is in the HF not the LF spectrum... of course that does not prevent an effect generated by LF from producing HF "distortions"...
_-_-bear