I'd say you stand to lose more in the name of phasing issues unless you are REALLY careful and understand the nuances of micing...
For bass -- I'd say not worth it, on the premise that when reproduced through speakers, the \"ummmmph\" in bass tone below 100hz or so is largely non-directional (the ear can't really pinpoint its location), and recording it in stereo comes at the risk of unevenly ducking other mid-range/high frequencies in your mix when it comes time for buss compression/limiting. While the majority of the \"tone\" of a bass weighs in aroun 200-500hz, and these are direcitonal, I'd say the extra \"width\" in soundfield you may gain will more than likely just clutter the mix, instead of enhancing it. Maybe a real sparse arrangement could benefit, but I've never done it.
For guitar -- yes and no. If ya' just throw mics up without having a real critical monitoring situation, you could run into phase issues and the like. Again the extra \"width\" in the soundfield may just clutter the mix. But... if you're real careful and have a nice room tone, you may be able to do some interesting things with space and dimension. There's a great little section on the \"Haas Effect\" in bob Katz's book that explains what it is, and how two mics can be used to capture reflections in such a way that you can create a true stereo image in a very natural manner. Note this isn't for use in a little vocal booth -- you really need a good room with pleasing ambience to benefit. Another option is two record two channels, though not necessarily for stereo use. You can mix the two mics to a mono channel, and by adjusting each one's level and position change the tone drastically. Sometimes \"phasing\" issues will work for the better in this scenario, acting like a natural eq filter. But again -- more times than not this is a long road filled with trial and error, so give it a shot, but don't expect it to sound better just cause you're throwing a second mic at it...
The same rule applies for vocals. If it was a sparse vocal or chior arrangement, and you have a nice room, and want a very roomy laid back sound... maybe it would work. I've never had a vocal situation that warranted it, and therefore never tried. More times than not, you want the proximity effect that a single close mic can deliver, and you want minimal room tone so you can compress freely without bringing up the natural ambience (this want you can add the flavor you want later.) So I'm thinkin when recording your average, dense Pop music arrangement, record \"lead\" or up-front sounds in mono 99% of the time. When it comes to little classical, bluegrass or jazz groups, or very sparse vibey arrangements (like piano and vocal) -- that's where I'd try puttin stereo micing techniques to work (if you have a nice sounding room.)