The type of reasoning put forth by Stehno is flawed and ultimately wrong. He incorrectly focuses upon the playback side of the equation. The real question is whether 2 channels of audio is sufficient to record the sound of an instrument playing in a real space? As a practical matter it's done all the time and with good results, but it's not perfectly done. There are at least 10 different microphone techniques for two channel stereo recording (spaced omni, X/Y, Blumlein, Jecklin, binaural, etc.) and that's a tip off to the problem. There are too many techniques. If one of the techniques really worked as opposed to just doing an okay job, then there wouldn't be a need for such a variety. Everyone would just use the one recording technique that worked perfectly. Now if you can't capture the sound of an instrument playing in a real space with only two channels, then it stands to reason that a two channel playback of the same is faulty.
If you're willing to settle for a rough approximation, let's use Stehno's 15-20%, of the sound of an instrument playing in a hall, then 2 channel playback is a proven technology that works well. Personally, I'm more than satisfied with just 2 channels. However, if you really want to fully capture and reproduce all the sound being produce in a hall, you'll need more than 2 channels.
If you're willing to settle for a rough approximation, let's use Stehno's 15-20%, of the sound of an instrument playing in a hall, then 2 channel playback is a proven technology that works well. Personally, I'm more than satisfied with just 2 channels. However, if you really want to fully capture and reproduce all the sound being produce in a hall, you'll need more than 2 channels.