Hi Craig,
Nope, you are entirely lucid. It makes no sense whatsoever, and produces no audible benefits, other than lowering total resistance in the cables to the speakers. But you can accomplish exactly the same thing using biwiring or a thicker gauge of single cables. Of course, if someone goes to all this trouble, his or her active imagination may stimulate notions that there are audible differences when none exist. They have the expectation of an improvement, so without experimental controls in place (an A/B/X comparator or double-blind setup with instantaneous level-controlled comparisons), they'll likely hear an improvement.
Regards,
