Wow, now you're preaching at me about what is audible. Until there's an error spectrum available, let's not even bother with that, ok?
.2dB difference can JUST BARELY be detected in a time-proximate ABX test. Barely. By trained listeners. If you want a citation, I cite me.
If the cables have enough resistance to interact with the headphones, something is wrong, either you shouldn't be using those cables, or those headphones, or something of that sort.
You need to keep the headphone position stable, one way or another.
Facing another direction may or may not be sufficient, depending on many more things that I care to go into right now.
I think you have no idea how good the human brain is at incorporating all senses even when that's not intended. Nobody, in fact, has ever shown they can AVOID that.