As long as the insualated wheels exist on all axles it should not matter which position they occurpy.. However,
there could be a breakdown of the insulation that is permitting a degree of conductivity, thus the 'short' effect.
The position of the wheels you describe is usual for cars with lighting. Are there power
pickup brushes against the rear of the wheels? If so, they could be worn and intermittently shorting.
One other thought...have the axles of this car been painted with 'conductive' paint so it would trigger
You could test the insulation with your multimeter set to Ohms. One probe on the wheel, the other on
the axle. Any reading would indicate a leakage thru the insulator. That pair of wheels should be replaced.
Metal car body, metal truck frames, metal axle... yes it absolutely matters.
If any of those components are plastic it doesn't matter, but with everything metal the insulated wheels absolutely have to all be on the same side as the metal wheel-axle-truck-body completes a circuit across the tracks.
This was an old Athearn tinclad boxcar and I had to shim one side because the coupler was hitting the crossover. Apparently when I did that I rotated that truck 180 degrees and the underframe is tin so... The crossover is relatively new and I get sudden stops on a regular basis as I find out which couplers are too low.