I don't know about you people, but I have only two analog inputs in my brain.
A strong argument for headphones (and binarual recording), but not a good one for recordings made for playback over speakers.
If I am in a very nice sounding room and I want to hear it I will use a sub-cardioid mic. If the room doesn't sound good I go for a more direct sound with a hypercardiod. I would make the argument you can get all the room sound you want with a single pair of mics. (assuming a standard 2 channel PA)
Totally reasonable and true. No need to assume a standard PA, non-standard PA or no PA though. You can get all the room sound you want with just one mic for mono, two for stereo.
Here's an interesting thing about that though- a mono recording is much less tolerant of the level of reverberant room sound than a stereo recording. The the sound becomes muddled, unclear and objectionable much quicker as reverberant levels rise compared to direct sound levels when there is only one microphone and speaker, compared to two. The same trend holds true moving to additional discrete playback channel counts. So in that sense (admittedly one of the few), a decent mono recording is actually harder to make from an overly reverberant room than a decent stereo recording, and that trend continues as the channel count increases. Not much else with all this get's easier, but that important aspect does.
An important part of finding optimal microphone arrangements for multi-channel recording is increasing the directivity of the three front channels and the separation between them over what is provided by optimal 2-channel stereo configurations. Again, that same principle applies to three (or four, or more) channels which will be mixed down to stereo, which is a pretty common thing around here. However, once mixed, that extra tolerance against additional reverberance I mentioned above is lost. That aspect is entirely dependent on increasing the number of playback channels and maintaining that increased degree of separation that was provided by an appropriate microphone configuration all the way from the microphone to the individual speaker.