First of all, don't believe everything you read in trade rags. When in doubt, use your ears. Listen to the finished track in question and ask yourself how many separate guitars you actually hear in the mix. The answer will usually be 1. If that producer needed 20 tracks of mics and alternate takes to get to that 1 guitar part in the final mix, then so be it. But don't look at those 20 tracks and think you're hearing 20 different guitar parts because generally you're not hearing more than 1 or 2.
 
Beyond that, everyone else has already given you great advice. Treat reverb and delay as though the mix knob controls the price and be careful about how much you spend. Also if the song really does contain multiple unique guitar parts, hard pan them as much as possible. Rhythm guy? You get the left speaker. Lead guy? you're in the right speaker. Singer guy who thinks the the other two guitars aren't enough and the world needs his 6 string contributions on top of it all? You get panned 1/3 one side or the other depending on what you're playing compared to what they're playing. And then you get eq'd and compressed until there's no danger of anyone getting hurt.

I think think the voicing of the separate parts are important. Don't let them step on each other. Yse different inversions.

If one guitar is playing an "open C" chord, then have the other guitar play a three note chord on the top strings at the eighth fret, for example.

If there is a keyboard part, make sure it is playing in a different register to the guitars, etc etc..

I remember watching a video called "Mix it like a record" this help me alot concerning various delays and reverb buses with in a mix......just me thinking out loud again.

JohanSebatianGremlin
First of all, don't believe everything you read in trade rags. When in doubt, use your ears. Listen to the finished track in question and ask yourself how many separate guitars you actually hear in the mix. The answer will usually be 1. If that producer needed 20 tracks of mics and alternate takes to get to that 1 guitar part in the final mix, then so be it. But don't look at those 20 tracks and think you're hearing 20 different guitar parts because generally you're not hearing more than 1 or 2.
 
Beyond that, everyone else has already given you great advice. Treat reverb and delay as though the mix knob controls the price and be careful about how much you spend. Also if the song really does contain multiple unique guitar parts, hard pan them as much as possible. Rhythm guy? You get the left speaker. Lead guy? you're in the right speaker. Singer guy who thinks the the other two guitars aren't enough and the world needs his 6 string contributions on top of it all? You get panned 1/3 one side or the other depending on what you're playing compared to what they're playing. And then you get eq'd and compressed until there's no danger of anyone getting hurt.

 
These are really good words!!  If you listen to most recordings you rarely hear more than three instruments playing while the vocal is happening (usually drums/bass/some type of rhythm instrument).  There are exceptions, of course, but it's a good general truth to follow. 

A picture is worth a thousand words, and all that...this shows how in many places in a song, there are only a few tracks playing. The overall sound is still big, though, because they create so much space for each other. (Of course this is only one song, but hopefully you'll find the dissection helpful.) 
 

 
This is a recent song. The top three tracks - Drums, Rapture Pro Gibson Bass, and Rhythm Guitar play pretty much throughout the song (which is rare for me - normally I'll drop these out in a few places).
 
The green tracks are vocals and the lower four tracks are all Rapture Pro parts. Note that until measure 26, most of the time there's a maximum of four tracks playing. There are two short parts where in one the harmony vocal plays, and in another, there's a drone. Still, that's only five tracks.
 
Starting at measure 26 is the "big chorus." The Staccato Key and Plucked Key are panned oppositely, and work with the guitar to provide texture. Even the "biggest" part of the song so far is 7 tracks.
 
The solo starting at measure 42 has only 6 tracks, and note how soft the guitar track is - that's to make room for the lead at the bottom. The "massed vox" are just a choir where I recorded a bunch of parts into Take Lanes, then collapsed them all to make a chorus. They provide a background.
 
8 tracks play only at the very end, where the two vocals are both happening. The Power Chords in track 5 are there to add some final emphasis.  

Two things that I've found make a difference, are:
 
1) Turn down the distortion on guitar tracks when using amp sims. Unless you're going for a full-on metal sound, you don't need all that distortion. Yes, it sounds good when jamming or tracking. But you'll soon realize that in the mix, you do not need all that distortion. I was really surprised at how little distortion you actually need to make guitars sound great in the mix. Lately, I've been using the old Fender amp models on Amplitube (Twin Reverb, etc), and adding an overdrive pedal, with just a little bit of drive for rock rhythm parts. Then I also have a completely clean rhythm guitar with a chorus pedal. When all the guitars come together in the mix, you'll realize how little distortion you actually need.
 
2) The capo is your friend. When tracking multiple guitars, try using a capo on one or two of them, while playing the others w/o the capo. Same chords, just higher up the fretboard. For instance, record a clean guitar track just playing normal open chords or bar chords. Then record a dirty track, same chords, but with the capo. This gives you a really nice full sound with minimal guitar tracks. And definitely put the two on different amps to differentiate them.

Just wondering why this thread is not in techniques?
Good stuff everyone.
Lot of tips I've always used.
.
I'm not a fan of doubling vocals,, always sounded cheesy to me. But 5 or 6 guitars parts is not hard to do if you are creative with how each part is recorded and using different  Guitar voicing's, different guitars, different amps. panning

I've been thinking more about this since the other day. I've got another thought of it that some may find... well... perhaps a bit rude. So forgive me in advance.

Way back when I was still in my teens, I had a girlfriend who had hair that was about shoulder length. She really wanted to grow it longer but it always seemed a struggle for her to get it to grow much longer than it was. One day while she was sitting the chair at the salon, she asked the stylist why she had so much trouble getting her hair to grow any longer than it was. 

Her stylist didn't pull any punches. He said if you want to grow your hair longer, stop cutting it. Should have been obvious, but it wasn't. Conventional wisdom at the time (maybe still?) said if you want your hair to grow faster, cut off the split ends. But the reality is, if you want it longer, stop making it shorter. He was right. She stopped allowing anyone to cut of her hair no matter how small or insignificant and less than a year later, she had the length she'd been looking for.

Ok so now the rude part. If you want your mixes to be more open, stop making recordings that are less open. There are a million and one tricks to adding multiple guitar parts to a song and ending up with a clean mix that isn't terribly muddy. But not terribly muddy isn't necessarily the same as what one would call open.

Maybe I'm getting hung up on semantics here and maybe my definition of an open mix is different than someone else's definition. To me, an open sounding mix has space. It has room to breath. And every element can easily be picked out and clearly heard. It is very difficult to create a mix that has space and room to breath when the arrangement itself is cluttered. 

Leaving open space in the mix requires... well, open space. Mixing 101 tells us that one of the primary jobs of EQ in a mix is to carve out a unique sonic space for every element of the song. The more elements the song has, the more frequencies will need to be carved into their own unique space. Which means each frequency space will therefore have to be smaller and therefore less spacious. 
 
If you want something that is open and has space. But the parts that are there aren't seeming like they're enough and you feel you need extra guitar parts happening, maybe its time to rethink some the parts that are already there. Maybe some of them can come out? Maybe something already there just needs a bit more flourish? I don't want to get too deep into the weeds on the producer side but the point is, perhaps the best solution to this perceived mixing problem is to treat it as a producing/arranging problem. Just a thought.

Like I have said before if many of you suddenly had the opportunity to work with a really great producer and you showed them where you were up to the first thing they would do is rip out about 60 percent of the stuff that is in there now.  They would be saying do you need this, do you need that, why have you got 15 acoustic parts here when only one will do.
 
Listen to Steely Dan for the ultimate in economical parts.  The trouble is the music in most cases is boring and not great so people are trying to make it better by filling it all up but it just does not work.  The music say in Steely Dan's case is brilliant so then what is going on there is just the minimum to get the message across.  I like Craig's example above.  He has sort of nailed it actually.  I bet that music sounds good and big but still has the message with only 7 or 8 tracks.
 
You should be able to see what I call the black backdrop behind the music.  If you cannot then you are going down the wrong path.  What you guys are doing by piling it all on is creating what I call a grey backdrop.
 
No amount of carving out frequencies and stuff for parts in other parts compares to not having so much going on in any one point.  All fantastic mixes have the back backdrop in common.  Even Dark Side of the Moon does.  It is so obvious to me.  You can walk around inside the mix and touch it.
 
The trouble is we have too many tracks available now.  There is an interesting thread going on in the Techniques area on why working with very limited resources makes you better.  I only had 4 tracks to start with (in 1980) but spent years perfecting the art of making incredible (electronic) music with just those 4 tracks.  Every sound counted and had to have a god damn good reason for being there.  I then moved up to 8 tracks for years and really mastered the art of making great music with only 8 tracks.
 
When I studied my Jazz degree many years ago it always used to blow my mind how complex the music could get with only 3 or 4 musicians.  You should be able to do it with same number of tracks.  When Return to Forever came out here only a few years ago there were not many of them on the stage you know but Oh My God did the music get complex.  Why?  Well for a start the music was ridiculous and so was the musicianship.  Seems that the better these things get the less you actually need to get the message across.
 
Even when you want to create a dense complex piece it is still amazing how little you need to actually do it.  In complex classical music yes with 80 musicians or more you can still see the black backdrop behind.  Because many of them are playing the same parts and there is not actually not so much going on when you listen right into it. 

Jeff Evans
Listen to Steely Dan for the ultimate in economical parts.  The trouble is the music in most cases is boring and not great so people are trying to make it better by filling it all up but it just does not work.The music say in Steely Dan's case is brilliant so then what is going on there is just the minimum to get the message across.  I like Craig's example above.  He has sort of nailed it actually.  I bet that music sounds good and big but still has the message with only 7 or 8 tracks.

All your comments are spot on...but with respect to the last quoted sentence, at the risk of great personal embarrassment I did a reference mix so people could hear for themselves what that number of tracks sounds like. Below is the **UNLISTED** link to the video, please don't share it...the song is only a day old so it's very rough, and I'll remove the link before too long to avoid additional personal embarrassment. If you want to jump to the "interesting parts" starting at measure 26, they're about a minute into the song.
 
On the plus side, it is to my knowledge the only song in the known universe where the words reference "To Know Him Is to Love Him," a Phil Spector song performed by The Teddy Bears