John,
My reply was not meant to preclude Duke's or anyone else from replying. I appreciate your interest and the question. I saw you addressed it to Duke and apologize if my reply was taken the wrong way.
Sorry I somehow missed this post of yours, John.
I don't have much to add to Jim's answer, but here's my perspective:
My first-generation bipolar bounced the rear energy off the wall behind the speakers, and my second generation bounced it off the side wall. Both called for generous spacing between speaker and wall.
According to Toole, the theoretically ideal direction for reflections, from an envelopment-enhancing standpoint, would be about 60 degrees to either side of dead center (like, 10:00 and 2:00). But we have to balance that against the amount of path-length-induced time delay - generally speaking, the longer the better, and as you pointed out the ceiling path length is indeed longer. Jim tested these two competing formats against one another and found that the greater time delay with the late ceiling splash was the clear winner, and having heard his system, I agree with him.
I would not have thought to try Jim's late ceiling splash configuration on my own. I would have relied on my "little knowledge" (that dangerous thing) and dismissed it as obviously inferior based on my armchair analysis.
There may well be more going on than just this longer path length. I don't think Toole looked at the effects of spreading out that first reflection somewhat, and I think we get a bit more of that with the ceiling bounce, along with that nice long time delay.
