Hi There;
It stands to reason that the more you do to a signal, the lower the purity at the output. In the analog domain, this is certainly true, especially with crossovers, which cause phase errors. However, with a DSP engine, it's easy to create complex low and high pass filters that have no phase error. Also, the only summing that is involved is in the sub output, and this doesn't result in either an increase in noise or dynamic range. Once a signal is summed, in the analog or digital domain, it can't be 'split apart' again, and there isn't a reason to do this in a home theater applicaton.
The only issue I have with running speakers as 'small' is locational phase error, resulting from a speakers' bass being redirected to another place in the room. Your ears can't discern directionality below a certain frequency, but there is potential for differences in arrival times, which messes with inteligibility, etc. In practical situations, though, I haven't found it to be a real problem.
On the other hand, in my personal experience, I find that setting the center channel to 'small' helps keep movie dialog inteligible, as long as the crossover frequency is set well below the dialog range. This probably wouldn't be an issue with a big full range center speaker, but I don't have one, and likely never will.
My recommendation is that, if budget and space allow, you can use full range speakers all the way around and avoid the redirection issue. However, if this is a theater, you have to keep in mind that movie soundtracks are harder on speakers than music, unless your musical taste includes explosions and such. You would want speakers that are capable of sub-30 Hz performance, in my opinion, or you risk damaging them.
Mike
