If I were using two mics that I couldn't get coincident, I would be using an odd-interger rule. Yeah, there's going to be phase issues, but so long as you choose an appropriate intergral distance, then you'll be fine.
To be totally honest, something as vauge-sounding as a "3-1" rule is more than likely to have ambiguosity in its interpretation.
My source:
http://humbuckermusic.com/acguitrectec.html
Sure, it's a webpage, but, hey, this is how ambiguity happens. If you can show me a textbook source that defines "3-1" rule as a mutlipe-mic technique for different sources, then I'll be leading the campaign to burn this site down.
However, I'll think you'll find that, logically, the 3-1 rule is something that can apply both to phase cancellation from a single source AND noise rejection from other sources that you're recording at the same time.
Just apply a bit of brainpower here- why would something ambigious show up in both examples?
It's because the examples are linked.
When you're talking about noise rejection, it's because the drop in power over triple the distance is a factor of 9, which, by my dodgy 0300 maths means an approximate drop of something like 10dB. Also, 3-1 is kind of a nice number, they're both part of the fibbonacchi (i can never spell that right) series, and a seperation of 3-1 isn't going to take up a massive amount of space.
The 3-1 rule in realtion to phase cancellation is simple- when multiple-micing an instrument, as in the top picture on the above link, you're going to want different signals going into the mics, but you're going to want them in phase, right?
The "1" part of the rule keeps both signals in phase, and the "3" part deals with the noise rejection from what are, essentially, different sources.
So, how's about we lay this argument over symantics to rest, and help the OP with his question about the phase reverse switch? As I said in a post somewhere else, you should consider reading a bit into wave theory and what exactly you're doing when you're reversing phase etc.
This thread should give you good grounding into the basis of how to keep things in phase, but you should at least look up the superposition of waves and basic phase shifts somewhere.
http://en.wikipedia.org/wiki/Phase_(waves)
that might be a good place to start. It also shows the proof behind Timothy Lawler's point #4- a shift ina tme is the same as a phase shift.