WC,
As thezaks (Dave) mentioned, another option for large soundstages is any tall line source speaker. The GT Audio without the subwoofer addition of faxer is an example. It is tall and narrow. Another example is one of the models of Wisdom Audio which makes SOTA ribbons. Both of these speakers have more accuracy than any Maggie to my ears. These tall speakers give a wide and deep stage, but with less bloating and smearing than large panels. A singer has the dimensions of a baby elephant compared to a big elephant from large panels. The reason is that the large panel has duplication of drivers from all its parts, so is like a dynamic speaker with dozens of little drivers spread on its large blanket. This is NOT true to life. Imagine 12 musicians spread out in a horizontal line across a stage. The whole stage is wide, but each musician has small width and is more like a point source. The point source speaker will most accurately reproduce the true point source of each musician. The line source is less accurate than the point source, because it has vertical duplication of the drivers, but is needed to provide more volume output from the inefficient transducers. Least accurate is the big panel. The big panel is only appropriate for reproducing very large instruments like the 9 foot piano or church pipe organ.
A horn speaker is an interesting case of a hugely dynamic speaker as a point source. The small driver in the throat of the horn is naturally amplified by the horn, so you get the best dynamics of any speaker, better than the Neo or Soundlab or huge Maggie 30.7. Images are more focused from the horn than from the large panel. The Klipschorn is a point source in the midrange and HF with a large wavefront launch in the bass only. This imaging is true to life, because mid/HF instruments like acoustic guitars, voices and brass are near-point sources, whereas bass instruments like the string bass are much larger. I remember you didn't like the "shouty" quality of some horn speakers you heard in the past, but the horn speaker really yields the most lifelike images. Here is where a dedicated EQ like the Rane can save the day by modifying the tone to your taste.
But I still say that the best compromise for you might be the Persona 9H. The bass controlled system might make the inferior aluminum driver as accurate or more so than the Be driver for the bass used in the model B. The big IF is whether the Be midrange driver equals or exceeds the resolution of the stat driver of the Neo, so I would like your opinion on that. It is about resolution, not bigness. If the 3F still doesn't give the midrange resolution (think clarity of words in the song) of the Neo, then you might as well stick with the Neo. In that case, the 9H will just be another expensive dynamic speaker that still can't compete in resolution to the electrostatic.
As thezaks (Dave) mentioned, another option for large soundstages is any tall line source speaker. The GT Audio without the subwoofer addition of faxer is an example. It is tall and narrow. Another example is one of the models of Wisdom Audio which makes SOTA ribbons. Both of these speakers have more accuracy than any Maggie to my ears. These tall speakers give a wide and deep stage, but with less bloating and smearing than large panels. A singer has the dimensions of a baby elephant compared to a big elephant from large panels. The reason is that the large panel has duplication of drivers from all its parts, so is like a dynamic speaker with dozens of little drivers spread on its large blanket. This is NOT true to life. Imagine 12 musicians spread out in a horizontal line across a stage. The whole stage is wide, but each musician has small width and is more like a point source. The point source speaker will most accurately reproduce the true point source of each musician. The line source is less accurate than the point source, because it has vertical duplication of the drivers, but is needed to provide more volume output from the inefficient transducers. Least accurate is the big panel. The big panel is only appropriate for reproducing very large instruments like the 9 foot piano or church pipe organ.
A horn speaker is an interesting case of a hugely dynamic speaker as a point source. The small driver in the throat of the horn is naturally amplified by the horn, so you get the best dynamics of any speaker, better than the Neo or Soundlab or huge Maggie 30.7. Images are more focused from the horn than from the large panel. The Klipschorn is a point source in the midrange and HF with a large wavefront launch in the bass only. This imaging is true to life, because mid/HF instruments like acoustic guitars, voices and brass are near-point sources, whereas bass instruments like the string bass are much larger. I remember you didn't like the "shouty" quality of some horn speakers you heard in the past, but the horn speaker really yields the most lifelike images. Here is where a dedicated EQ like the Rane can save the day by modifying the tone to your taste.
But I still say that the best compromise for you might be the Persona 9H. The bass controlled system might make the inferior aluminum driver as accurate or more so than the Be driver for the bass used in the model B. The big IF is whether the Be midrange driver equals or exceeds the resolution of the stat driver of the Neo, so I would like your opinion on that. It is about resolution, not bigness. If the 3F still doesn't give the midrange resolution (think clarity of words in the song) of the Neo, then you might as well stick with the Neo. In that case, the 9H will just be another expensive dynamic speaker that still can't compete in resolution to the electrostatic.