I've never performed measurements, but my ears have always detected phase problems when the rather ubiquitous Yamaha Piano Samples are involved (which all sound rather alike, they must have sampled the same model C or something).

At any rate, the stereo pianos seem to blend better -- and sound more like a realistic grand piano -- when the two channels are amplified separately as vs summing to mono.

The blending of the two signal aources in the compressible medium of air takes advantage of the physical property of elasticity here and I think that is a factor. I also think one could write a thesis on the subject which could get rather deep rather fast if all the possible ramifications are investigated.

At that point, it is cheaper to simply move to true stereo amplification.

BTW -- when onstage I prefer to have a separate stereo keyboard amp and speaker system, fullrange, but leave the PA front as it was. The piano when amplified through the typical PA front, separated extreme stage L and R as most are, doesn't sound right to my ears either, unless you want your audience to perceive themselves sitting inside the world's largest grand piano, a listening situation of the absurd. Of course, the Bose Compact situation we are talking about can neatly sidestep that problem entirely as inherent in its design nature -- if you use two in stereo. I think the Compact, with its multiple drivers, does indeed impart a slight bit of its own phase situation, but that is part and parcel of what the design yields and certainly not detrimental for the live performance in the right sized venue. Of course I'd never use that system for audio reference or monitoring for recording, though, that'd be the kiss of audio death and a task for which the system was never designed nor intended to do.

--Mac