For those who are confused about when to use mono tracks and when to use stereo tracks, this is a great video tutorial. The guy who does this is very good at explaining things. Even though it's geared towards electronic music, the stuff in here still applies to other genres. Well worth a watch. 
 

 
If like the style of this guy and his teaching then you might also enjoy his video on mixing sub bass, again a very in depth treatment and very easy to understand. 
 

 
The Point Blank Music School seems like a quality setup and I would definitely consider taking one of their courses. The guy in this video is the head of the school. I felt like I got a lot out of these two videos. 

Yes a good video.  I think one good point is that if you are dealing with several stereo sources at once. eg electronic music. Panning them all into the same positions can make them less clear as a result.  So you can alter the widths of several sources and this in itself is a form of expression within the mix.  I tend to make at least one very wide as well eg a slow moving pad or something.  Listen to them (in solo) before setting their widths because some are nearly mono and are worth putting into mono.  It is good to have a few that are about half as wide.  You can pan a narrow width around to the left and right within the overall image.  Narrower sounds are easier to place in some respects.
 
But what has been mentioned in these forums is recording stereo instruments especially hardware synths.  And wrong advice in that it is best to record from a synth in mono and then apply stereo effects later.  A better way is to always record any stereo instrument in stereo.  Often the images that are coming out are very nice and well worth hearing and tracking at least.  You can always alter them later and even bring them back to full mono again and apply some other stereo processing later.  Sometimes a little stereo processing over a stereo sound works great too.
 
There are times when you may want to use 5 stereo sources and pan them all into the exact same place as well.  Suppose you want to make a very lush pad using 5 layered synths all in stereo making a slightly different pad sound.  (eg 5 synths per note)  It is good to pan them all to the same place sometimes as it still gives the immense sense of lush but the layers are a little vague to  pinpoint which is an effect you may actually want.  You could put all the 5 stereo synths into mono and pan them in five places across the image from L to R.  That will sound different to the first option.  You may pinpoint individual layers too much in that situation.
 
Most VST's are going to be in stereo too so it well worth auditioning them carefully around the time you are doing your mix. Then is a good time to alter the original stereo imaging perspective to your mix perspective.
 
Even experienced mixers can forget how multiple stereo sources are being returnd to a mix so it is good to be vigilant on those things.

seems like most of it is just common sense,
IF you have spent a lot of time mixing, 
you will have already figured most of this out.
 
\IF you are just starting,
really good tutorial.

i'm finding that the longer i mix, the more i get away from stereo effects,
and use mono reverbs and delays, and pan them with the panning of the individual track.
 
 

Yeah most synth presets give you maximum everything - maximum width, maximum frequency range. You almost always have to chip away at them to make them fit.

Personally I've been leaning towards a more LCR approach of late and I've found that it's giving me a much clearer, much more well defined sound. I'll maybe have one stereo part in there, but will usually narrow its width by half or more. Those hard panned mono sounds just add loads of width to a mix. The stereo reverbs have also taken a back seat to mono reverbs, also panned LCR. I find having something panned one side and its mono reverb panned to the opposite side does a lot to add space. If I use a stereo reverb it's either when the track is very sparse, or I'm just adding a tiny barely perceptible amount to add a little cohesion. I'll often narrow a stereo reverb as well. No need to have them taking up the whole field.

I don't soft pan much now, unless it's something like a subtle vocal delay.

batsbrew
i'm finding that the longer i mix, the more i get away from stereo effects,
and use mono reverbs and delays, and pan them with the panning of the individual track.
 
 

Same here.
 
A little while back, I've even started building my balance in mono. In fact, I often forget that I hit the mono switch and keep working in mono until something brings it back to my attention, like a particular pan. Overall, I spend a lot of time working in mono.
 
Usually, if it works, it's only going to be better in stereo.
 
As for the rest, I tend to use the stereo tool in my DAW to constrain the stereo image of certain instruments (like overheads) to a certain width. I've been re-centering things quite a bit lately.
 
If I pan something all the way left or right, it's for effect or because I want it to really stand out. It's what seems to help with my mixes, though I can appreciate mixes done w/ a different approach.
 
 

I'll need to watch the video later as I can't see it where I am. Looks interesting.
 
 Some audio interfaces allow for  secondary adjustments of the output channel. One example would be my Presonus Firetube studio software mixer. The accompanying software mixer runs underneath Sonar X3. Everything that passes out of Sonar also passes into and out of that software mixer, so  both level and panning are affected. I can take my mains out from the Presonus mixer and pan each channel either full L/R or keep each channel a center mono position.
 
You can do a lot of panning work and loose it if that secondary mixer isn't set the way you want it.