And even where it's not, 'stereo' playback simply cannot recreate real space.
I agree that two-channel playback cannot recreate a realistic acoustic image of the whole scenario of musicians, concert venue reflections and perception of envelopment at the same time. People trying to push the reproduction of stereo recordings towards a more ´realistic, wide´, ´holographic´ or enveloping impression, they inevitably will compromise on other imaging aspects.
What it can do with recordings containing such information, is creating a plausible image of the event, balancing phantom source localization, stability thereof, depth-of-field, adequate proximity and perception of the venue´s size as well as shape and absorption grade.
Surely it is of no matter whether the spatial quality is real or artificial if it can be reproduced.
With spatial qualities on the recording which are not covering any real event (like an acoustic performance in a concert venue), it is actually difficult to, as a listener, judge what sounds right and what sounds wrong. I noticed many people expecting things like ´realistic ambience´ or ´singer has to sound like singing in the listening room´ kind of things from an electronic recording without meaningful reverb pattern. That will, in my understanding, lead to inevitable dissatisfaction.
I enjoy listening to various genres of popular music, but i always do so with the expectation that imaging on these recordings is something artificial. If I want to judge imaging of a reproduction system, I listen to classical recordings, preferably those which I know the original concert experience as well as the mixdown in progress of.
Holographic, great imaging, depth, width, height.....
I personally would avoid both the term ´holographic´ as well as the dimension of ´height´. The latter, at least with stereo recordings mixed for loudspeakers, cannot be reliable tracked to information which are on the recording, but is a rather random side effect of HRTF-related tonal imbalances.
´Holographic´ in my understanding is used by too many people in contradictive ways. Some use it to describe an overly ambient or coherent image of the phantom sources and reverb, others like very distant imaging, or very proximate one, with the most commonly found definition being phantom sources appearing to be ´stable but in a vacuum´ with no enveloping reverb. That is how I understand for example the posting about KEF Q speakers, as these oftentimes produce such image (which I do not like at all).
Its not only depth, its also height. I've experienced overhead sounds with just two speakers. I've also experienced sound behind my head. Instruments being played by someone sitting on the couch to my left or even sound coming through the wall to my right. I've had people come and didn't tell them what to expect...but they heard it too.
As mentioned, localization with no intraaural differences, is a real thing, but it is vastly based on HRTF-related tonal changes (see Blauert´s theories on preferred directions of frequency bands). Boost certain bands (like 800-2,000Hz) and attenuate the neighboring ones, compared to a reference source of the same timbre, and imaging will move to the back, same with 2,000-5,000 and the front.
But that is not something that is recoded and mixed on recordings for loudspeakers. Headphone mixes, for example binaural ones made with a dummy head microphone, are an exception here, but rely on headphone reproduction.
The wonderful thing about speaker measurements provided by Amir & Erin is it finally allows us to link our audible perceptions to proper data.
In theory yes, but if we talk about localization, ambience and imaging, it is actually pretty difficult to link perceived characteristics with data precisely. And in part this data is not even existing.
Take localization precision/stability as an example. What contributes most to that, in my experience, is coherent position and size of the relevant sound sources (midrange drivers, tweeter), as well as interchannel amplitude deviations (which are oftentimes not measured as it requires several units of one and the same speaker on the bench) and tonal coherence of early reflections (side wall, ceiling, floor) as well as in the listening window. Partly, traces of these characteristics can be found in spinorama, but the data is far from sufficient to really predict anything but very clear deviations from the ideal, IMHO.
The second aspect is that directivity and resulting reflections/reverb in the listening room, seemingly are not really understood in their implications on imaging and ambience. At least, that would be my explanation for so many speakers being lauded for excellent measurements, when they actually show a pretty problematic behavior hinting to compromised imaging, such as narrowing directivity towards higher frequencies, or significant lobing in the early reflection windows.