Other than sighted listening tests between the two units, how did you discover this?
Steering and panning cues are embedded in the original sound codec and spatially located by the Mixer when he created the track - are you asking us to believe that Unit A will 'interpret' DTS HD MA 5.1 somehow differently to Unit B and will therefore reproduce sounds at different spatial locations?
I am not sure how anyone can do these tests blind when you have to swap like a tonne of interconnects (11.3) between the two.
If listening tests aren't done blind, level matched to +/- 0.5dB, with instantaneous switching, they have no value. Well, they're not 'tests', they're just subjective opinion and as such are subject to all the effects of expectation bias, placebo etc etc.
As for the interconnect issue, I have mentioned this before. When I was using cheap optical interconnects and then swapping them to pricier ones - no difference with DTS/DD bitstream but there was an improvement with PCM data.
Not according to all the known science. Again, without objective tests, there's no point discussing it. You believe you heard a difference and I am sure you did hear a difference. But the difference was caused by the flawed 'test' method used (no level matching, no instantaneous switching, sighted) and not by the cables.