I am sorry Mapman but you are mistaken. As Onhwy61 is saying, you are confusing acoustic reflections contained in the recorded signal with those reflections produced by playback in a room.
To reproduce the spatial information recorded by the sound engineers it is not required to have reflections in the room. In fact the exact opposite is true, i.e. the less artefacts your room is inducing the better you will be able to hear the spatial information recorded on a disc.
Ill make an analogy with the bass (which is also affected significantly by the room) to explain why you are wrong when you write:
So the reflected sound of teh room during playback is needed in order to attempt to best reproduce what might have been heard live rather than just what is in a 2 channel stereo recording.
Of course playback room acoustics will be different than what existed during recording, so the two may never be exactly the same, but can come pretty close when both are similar.
What you are saying is equivalent to saying that in order to have good bass one needs to excite the modes of his/her room, and in order to obtain the best bass performance one need to excite in his/her room the same modes that have been excited in the room where the recording was made. Of course, that is not true. To hear the most accurate bass possible in ones room, one has to optimise the position of the speakers and listening chair so no major constructive and/or distractive interference occurs at the listening position. (Assuming that ones speakers are capable to properly reproduce the bass on the recording, the best performance is achieved when one succeeds to completely eliminate the artefacts induced by the room - which of course is not really possible).
When it comes to soundstage and stereo image, the room induces even more artefacts than in the case of bass. For example it can: 1) make a bigger than life soundstage, 2) kill the soundstage completely, 3) make a 1-, 2- or 3-dimensional sound stage, 4) make the size of the instruments bigger than in the real life, 5) kill the stereo image, 6) shift the stereo image to the left or right, etc.
Clearly, very many things can go wrong because of the room and the best way to avoid all these problems is to try to "get rig" of the room. if possible try the following experiment:
First listen a recording with rich spatial information on a good pair of headphones. You will clearly hear a large soundstage and pinpoint stereo image. (You may like the same recording via your speakers better because it has a large sound stage but that is an artefact induced by the room. Of course, there is no problem with that, many people like the bass reinforcement caused by the room which can be beneficial sometimes.)
then listen the same recording via:
1) A pair of monitors in "near-field" mode with the monitors situated as far as possible far from any room boundaries. (This is what the Cardas method is trying to achieve.)
or
3) A pair of horns by sitting not too far from them, i.e. far enough that all drives are well integrated.
I am very sure you will end up with very similar results, i.e. a beautiful and well defined soundstage and stereo image that have been produced with no (or minimal) secondary reflections.
Finally, regarding your experience with the MBL system, I can imagine that under well controlled conditions and with the right recording they can create a wonderful effect as our ears like to hear (or better said are used to hear) reflections. However, I doubt very much that the MBLs are able to pull that trick or to sound accurate in every room with every recording and with every type of music.