With the andecdotal result: bandlimited to 250Hz second order Q=0,5 (aka LR) the remaining sound of my stereo isn't "roomy". Room ... whatever You name it needs seemingly at least some mids to materialize...
In another thread apparently you described your system in more detail:
Practically I've got speakers with a directivity index of about 6(dB) @ 300Hz gradually narrowing down to about 9(dB) @ 10khz. These are toed-in by 45°. Due to the actual radiation pattern this alters the effective directivity index to, say 5 and 7 respectively. A DML with 'chaotic' phase rendition but very wide dispersion is added behind the baseline of the former two. Since the DML is a dipole, it was possible to minimize the output towards the listener. It was integrated by feeding it with a summed left+right signal, crossed over @ 1kHz with a quite shallow 6dB/oct. The level (measured in the reverberation filed), referred to the mains (both) was not critical at all. The difference between -10dB up to 0dB was gradual, not game changing. Don't say this is the holy grail, but it catches my interest still. Visitors like it, and my cat too.
So if I understand correctly, you have narrow-pattern main speakers toed-in by 45 degrees, and then behind your main speakers is a single dipole speaker facing the side walls (nulls towards the sweet spot and towards the "front" wall), which receives a summed (L+R) signal. Is this correct?
If so, your observation that the sound of your stereo isn't "roomy" makes sense to me. Imo your configuration is doing at least two things which tip the spatial perception away from the "small room signature" of the playback room and towards the "venue signature" on the recording:
1. The earliest reflections are the one which most strongly convey "small room signature", and your narrow-pattern main speaker with 45 degrees of toe-in minimize the amount of energy in the early reflections. Aiming the null of your supplemental dipole speaker towards the listening area prevents it from adding significantly to the early reflections.
2. The later reflections do not strongly convey small-room signature, but they do act as the carriers for the venue cues on the recording, in particular the reverberation tails. Because of the relatively late-arriving contribution of your dipole speaker, you end up with MORE beneficial, late-onset, spectrally-correct reflection energy than would normally be the case. This has the net result of strengthening the presentation of the venue cues on the recording.
So imo your configuration results in relatively weak "small room signature" cues from your playback room, combined with relatively strong "recording venue signature" cues, which makes it more likely that the latter will dominate the listener's perception. My guess is that the soundstage you hear varies a lot from one recording to the next, being dominated by the recording rather than by the playback room.
And there is one other thing your configuration probably does well: The contribution of your dipole speaker probably corrects the spectral balance of the reflection field, so that it more closely resembles the spectral balance of the direct sound. This would show up as improved timbre and reduced listening fatigue.
In my opinion.