One book-length post deserves another...
First let me differentiate between "rear" and "surround" speakers in regards to a 7.1 setup. I think, from what I've seen on this board, that the opinion of the majority is:
Diffuse radiating for surrounds, direct for rears. I don't personally have a 7.1 system, what I do have is a sub-ideal room layout with a 5.1 system.
So I will speak to the case of 5.1 with the surrounds being what this author calls rears. I previously had directional surrounds, when I got the QS8s I immediately felt that when watching movies the surround effects were more enveloping. An example of how this "enveloping" manifested itself:
Imagine a movie scene where the main character is moving through a jungle and some bird overhead is being it's usual noisy bird self. There is something akin to this in many many movies so I'm not going to bother with any specific one. With my directional speakers, birds of that sort always sounded essentially like they were coming from a particular region of the room between the surround and corresponding front speaker depending on how the effect was done. With the QS8s, that region gained "height" significatly as more of the sound ultimately reached my ears after bouncing from the ceiling.
When he says that you should place your surrounds as mains and listen in order to see how much the sound degrades, he is completely missing the mark. Most of the effects in movies used to give a "surrounding" sense are ambient noises that are not naturally caused by directionally radiating sources. It makes little sense then to send them to directionally radiating speakers.
Now then for music:
In stereo listening the recording engineer attempts to present the sound in a way that acurately represents some arrangement of musicians on a stage in from of the listener. This is inherently a directional source, and I don't think anyone would realistically argure that you should place non-directional speakers as your mains. When you bring surround speakers into the mix however, they are often used to represent the acoustic effects of the recording space. Thus the sound engineer is trying to represent reflections of sound that would be reaching your ears from all directions by mixing them into a discrete set of channels. Does everyone in their house place those channels at the exact same locations with respect to the listener as they were originally placed? Was the engineer able to accurately represent the sound in the first place? He is ultimately right that you will not hear what the sound engineer hears. But will it sound worse? From my personal experience thus far it sounds significantly better.
