The idea would be to calibrate similar to how amps do it now, put the mic at the seating position and let the computer do the work. Could even upload results to AI and have it spit out the 'solution'. With multiple drivers in the sound bar pointed in different directions can certainly come close, even today.
I'd never say never, there's plenty of things we take for granted that didn't exist 10, 20, 30 years ago. Time will tell if this delivers as a complete solution or just languishes in its current state as a knockoff surround "better than nothing" option. But technology seems to have a way of compensating in ways unimaginable by us mere mortals.
I'm hardly an optimistic person by nature but when it comes to iteration improvements over time, well just look at a cell phone for example of what can be accomplished when there's willingness and widespread adoption.