Just to repeat myself more clearly, in general when you hear a difference between two things that are most likely identical, you are in fact responding to other cues.
There may be slight differences in startup times of a track, say. If something is poorly designed, there may be occasional glitches, which are essentially inaudible, but allow you to subconsciously identify one thing over another, and then apply "more airy soundstage" verbiage to it, when in fact 99% of the time the sound is identical.
The ethernet versus USB versus whatever, for example. If the DAC is properly designed, the actual sound produced will be identical. However, in the case of, say, ethernet, the DAC (being properly designed) may choose to buffer slightly more data to guard against the somewhat stochastic nature of packet delivery on ethernet, resulting in a subtle startup delay. If your DAC actually sounds different, that is, the signal produced for any randomly selected 10 second interval mid-track, is not identical between USB and ethernet, it is broken and should be returned. But it probably does not.
Eliminating these variables is fiendishly difficult and not worth it.
It doesn't matter what the cues actually are. If your experience of the music is better with ethernet, or golden cables, or 500 pound granite turntables, then you are indeed hearing better sound. This is because hearing is a construct of the mind, NOT because the actual acoustic stuff going on in the air is any different. Usually.
Sometimes there's a real difference, usually because something is busted or poorly made.
