Let's run a chain of thought -
a car is running on DC power with (-) being a common plane throughout the car. DC power does not induce currents in adjacent insulated wires (other than at the power up and power down cycles).
the electrical turbine in the car (alternator) produces an AC current that is being converted via bridge into DC power and then stabilized at a certain charging level (let's say, about 13-14VDC). The AC currents will induce AC currents in adjacent insulated wires. Longer the wire, more chances it will pick up induction, any loops in the wire will magnify the induction effect (think of charging circuitry in the keybob and the ignition coil). Chances that the ground plane is "infected" with the AC current are high. Moving and relocating the cable will reduce, diminish or eliminate the "whine" but there is no solid scientific back-up, as what and how to do this task exactly, other then trial-and-error method.
Introducing a shielded cabling would eliminate the induction into the wires protected by the shield that is grounded, unless the shield is picking up the induction from the ground itself. In this case, the shield is actually magnifying the "whine" as it is always next to the signal wires
.
Possibly, to avoid this, the shielding would need to be grounded via Ground Loop Isolator that will prevent the AC currents from entering the shield directly. This would provide electrical solution to the issue.
The induction phenomenon is of electromagnetic nature, so, naturally, in addtion to the electrical insulation via GLI, magnetic insulation may also prove helpful.
I would welcome comments, suggestions and solutions - just have a pity on old fellow, don't hurt me too much...