I posted this on a thread about horn speakers after reading Frank's post about tweaking electronics being the most important area for the "believability of the reproduction illusion"
The resolution to this conflict about room Vs electronics may well be explained by Auditory Scene Analysis as Frank said.
Fundamentally, we construct what we perceive as hearing. This construction is based on processing of & the interrelationships between elements of the soundfield. As I understand it, we can cope with & adjust for certain types of anomalies in the signals if these anomalies are consistent. In other words a room has a certain characteristic sound that is always the same & we subconsciously adjust for this characteristic. It doesn't interfere with the believability of the reproduction - after all, it's what we are used to on a day to day basis - listening to people speak in different acoustic spaces - & we've learned this & built mechanisms to adjust for it.
However, if there are anomalies created in the electronics that vary based on the signal being processed, then we are less able to adjust for these & they interfere with the believability of the sound. Such anomalies are not encountered in the world of non-electronic reproduction & therefore they are much more likely to confuse our auditory perception. So, elements like, jitter & noise modulation, which are all signal-dependent lead to much less believability than certain room nodes or room characteristics do.
I believe that the general principle applies - if a low-level anomaly is regular or predictable then, within reason, we internally adjust for it in our processing & it becomes relatively invisible to our auditory perception. That's not to say that if this anomaly is removed we don't notice it - we do! Its what makes this hobby so interesting & what makes continual improvement possible - the discovery of & removal of smaller & smaller issues.
I believe that fully grasping the concept that "what we hear" is actually more correctly stated as "what we process" will lead to a better resolution of such issues.
Edit: If we treat hearing, not as a human perception, but as a learning machine that processes signals in a particular way (processing that we don't yet fully grasp), then we will make better progress in our understanding & ultimately in what makes for a "better reproduction illusion". We are too constrained by the concept of our abilities to instantly recognise frequency & amplitude differences & not so invested in how we might be processing signals over longer periods of time & how some discrepancies between these longer term relationships of the signals can lead to "less believability". This leads to the usual forum fights.
I'm not saying that room treatments don't matter - it's just a case of how we process these aspects differently & how they are perceived in different ways