I, alas, did not get to experience The Wall, but the theory behind it makes sense to me; of course, a good engineer can EQ* almost any system to sound pretty good, but it is always harder with full-range cabs than what I usually worked with, which was 4-way - separate cabs for lows, low-mids, etc. It would seem to follow logically that isolating the "voices" in their own speakers would make it even easier.
Were I to regain enough hearing to trust myself with someone's sound again, and could build the system of my dreams, I'd do a hybrid; a soundman out front & old-school right & left stacks, but of pairs of full-range cabs. Get a board with a lot of subgroups & use the group outs to drive a pair of cabs for each (vocal group, drum group, etc.), plus 1 group for subwoofers (bass, keys, and kick in that one as well as their dedicated pairs).
*The most important tool a soundman has - besides his ears, of course - is a graphic equalizer; the heart of the system.
Peter (who does, in fact, realize that he's dating himself by talking about a graphic EQ instead automated computerized stuff)