I'm conflicted about if/when to convert stereo tracks to mono when preparing for mixing a song.
 
For example, I've seen many mixers mono split the drum overheads in order to be able to process and mix them better.
 
What do most people recommend?
 

1. Do you ALWAYS split any stereo track for the sake of more options? Does that lead to too much work in the mix process?
2. What about stereo synths, stereo recorded acoustic keys, etc?
3. What about a stereo track with a mono source like a snare and one mic?

 
I'm so confused as to best practices. I know nothings written in stone, but are there industry best practices most mixers follow?

Stereo tracks are usually best reserved for stereo microphone array sources like XY, ORTF, AB, MS etc stereo pairs or stereo instruments like synths.  They are designed to be hard left and right without adjustments.  I would personally never split them unless I was doing some manual MS tinkering.  All other spot mics and single mic sources are best kept to a single mono track.   A single mic recorded to a stereo track is just a waste of information and serves no purpose, unless, for example, you are working with a stereo plugin that would potentially crash with a mono source.  I've had that happen a few times. 

Like what rumley said, when I was younger I used to actually convert all my vocal mono tracks to stereo cuz for some reason I thought it was better quality, then I gotta stereo splitter for my interface to record in stereo.All it did was endupp running up my cpu power. Thats years in the past tho, Istill on veery rare ocassions convert to stereo cuz it seems to have a stereo imaging effect and help it spread out. Some might say it's just personally how I perceive it.But I think sometimes it works out. Although I very rarely do it now. I definitely don't consider myself to be any kind of industry professional tho.

I record everything on mono tracks... for 100% control of what I want to do with it later. If I get handed a stereo track, I split it into two mono tracks. Panning laws apply to mono tracks... it gets more complicated and requires plugins to properly pan a stereo track.
 
In the end though it really doesn't matter how you get there... if the song and the mix is good - it doesn't matter what path was taken.

Like many others, I prefer to mix mono tracks. Mono tracks make it easier to get a more satisfying stereophonic width in the overall mix.
 
This can be a challenge when you use synths and samples, which normally default to stereo. Sometimes, converting them to mono results in information being lost or a thinner sound due to phase cancellations. You therefore have to approach each source individually rather than applying any fixed rule. I'll fold most stereo sources to mono and listen for any tonal degradation; more often than not, there is none.
 
Except in special cases, there is little value to splitting a stereo track into dual mono. It does give you more panning options, but Channel Tools does that without the bother of splitting the track. I do, however, split piano much of the time so that I can process the left and right hand parts separately. 

Bloomin heck... Yeah, I've just checked one of my synth projects and I leave the VST instrument track as a stereo track. Which is daft, because at the moment I'm actually disabling most of a soft synth's "on board" effects and using things like Timeless, for instance, in the effects bin instead.
 
I'll have to check, but if I then change a soft synth track to mono, will a stereo effect like Timeless stick work as a stereo effect?
 
I'll have to check... BRB...

If the track was acoustically recorded with two microphones and I would leave it stereo, If you have a mono track for each microphone from a stereo recording, mor often than not I adjust their gain to match and route them to a bus and do my processing on the bus. And I do exactly what Bitflipper says about synths and samples.

I do a lot with samples that need to be stereo for the sound they have (recorded in a concert hall). However since too many big sound sources result in too little room in the stereo field I will often use waves S1 to shirk down the size of the sound. This cleans up the samples nicely and gives more room to mix while retaining most of the big concert hall sound. But generally anything that is mono I leave mono and anything stereo I leave stereo.

Woodyoflop
But I think sometimes it works out. Although I very rarely do it now.

Conclusion: It very rarely works out.

MBGantt
I do a lot with samples that need to be stereo for the sound they have (recorded in a concert hall). However since too many big sound sources result in too little room in the stereo field I will often use waves S1 to shirk down the size of the sound. This cleans us the samples nicely and gives more room to mix while retaining most of the big concert hall sound. But generally anything that is mono I leave mono and anything stereo I leave stereo.

i have used S-1 for this same kind of thing before..
 
but whenever i leave it a while, and come back to it,
and take the S-1 off,
i realize that whatever S-1 is doing, sounds like clownphuckery to me.
 
at first it sounds great,
but then when i take it off, i leave it off.