It seems that there is a belief that imaging is a separate quality of loudspeakers.
I think that if a speaker has relatively flat FR, somewhat uniform directivity and is setup properly in space, there is no reason that it will not image well or give a nice soundstage.
How wide that soundstage will be will depend on many things.
Now, why friend's loudspeaker images poorly is very hard to tell when we don't know what speaker it is.
Can OP share speaker brands that they compared?
BTW, I also had Thiel CS3.6, but only for 8 months or so. Those were praised in the press for their time aligned drivers and imaging.
In the sweet spot, imaging, true imaging, i.e. recreating what is on the source material, is a factor of on-axis frequency response, equal on-axis frequency response between both channels, equal phase response between channels up to about 1500Hz, lack of significant rapid phase shift from 200-1500Hz (more debatable), keeping distortion under control, speaker width (angle), speaker distance (if driver integration comes into play), and importantly, keeping the level of reflections low especially under 6KHz (but not <200Hz). ....... That won't guarantee you will get what the person making the recording intended, but you will get an accurate representation of the "image" that is in the recording.
Will you like it? No idea. I am not you. Many, probably a majority will give up a bit of imaging accuracy for wider perceived soundstage, i.e. sound outside the speakers, but that is artificial, so you need to accept that it is artificial. The greater the level of direct sound to reflected sound, the better the imaging accuracy, but at the expense of other qualities most people like.
Smooth off-axis response will widen the potential sweet-spot, and allow, if desired, less raking of the angle, again allowing a wider sweet spot. In an untreated or lightly treated room it will also reduce room response peaks and valleys, which will just sound bad, and also in an untreated or lightly treated room, will effect "imaging", or "sound-stage".
Most speakers without major flaws, operating in an area of lower distortion, equalized, in an absorptive room, will image accurately. When you move away from an absorptive room, and reflections come into play, and you need to balanced direct/room response for equalization, then things get much more difficult.
When you look at things like open-baffle, dipoles, and line arrays, they don't inherently image better, they either enhance a quality that people like, or they "fix" an issue in untreated rooms. Open-baffles tend to have less side-wall reflections. Dipoles as well. Then they add in euphonic reflections that people like, that sense of space, whether it was on the recording or not. Line arrays reduce ceiling and floor reflections and provide a closer to 1/R drop in response versus 1/R^2 drop in response from about 500Hz and up (for typical floor to ceiling). That helps with untreated floors and ceilings but also improved sizes of the sweet spot.
No magic, just science, and all relates directly to how we perceive location.