There's no doubt that great performances are vital, but it's also worth mentioning that many of the publicly available multitrack masters are actually stems.

For example, the original recording of Queen's 'Bohemian Rhapsody' includes 180 tracks, the majority of which were used for the vocals. To accommodate this, several 24 track machines were chained together and later mixed down to a 'more manageable' set of stems for mixing purposes. I know this is an extreme example, but many of the available 'multitrack masters' have been through a similar process.

Brian May goes into detail in the following video...
 
Mixing Bohemian Rhapsody    

noynekker
Arrangements that are really well conceived also mix themselves,

I think this is a really important point.  I have arranged for jazz bands and pop/funk bands for decades.  Through most of that time, my attention was mostly on the music theory and making the musical lines sound hip.  I spent relatively little time thinking about the "mix" per se.  I have been work with ("tinkering with" is probably more accurate) SONAR and related technologies for the past 8 years or so. 
 
After about 6 months I had an epiphany that mixing and arrangement are (or should be) almost the same thing.  Both are all about letting the music be heard in its best context.  I'm not a great sound engineer at this stage. But I am a 100% better arranger for having this engineering experience.
 
The best arrangers (people like Sammy Nestico, Dave Wolpe, and Mike Tomarro in the big band jazz space) and the best orchestrators (Maurice Ravel, Warren Barker, Danny Elfman et al) seem to understand these concepts intuitively, whereas people like John Williams tend to just pound the drums harder when they want more emotion.

Cactus Music
I don't print  effects to "tape" and never did even before DAW's. 
Sure, my guitar I'll run through my stomp boxes, but vocals as far I know are now and have always been recorded dry. 

Well, there are different shades of "dry." Although you didn't do things like print reverb with vocals (at least I didn't, and vocals were often one of the tracks for which you reserved your hardware processors during mixdown), you were typically using one of dozens of carefully chosen mics whose frequency response matched the vocalist best, running through a well-maintained tube preamp, and then going to the tape via a channel strip with just a touch of EQ and maybe a limiter to catch the very highest peaks "just in case"...with tape adding a hint of harmonic distortion as soon as the signal got over -20. Anyone who recorded to tape was processing the signal whether they wanted to or not.
 
As to the guitar example, "back in the day" the intention in the studio was often to capture the live sound of a band. A lot of those musicians had effects they used onstage, they brought them into the studio, and recorded the resulting sounds. So while technically speaking processing may not have been added from the studio's backline, the tracks were still being printed with processing. Also with something like drums, the room itself added processing, and engineers took advantage of that "effect" with mic placement. 
 
So the bottom line was that a lot of the work required to get the sounds people wanted occurred before the signal hit the tape, and while the signal hit the tape. Although it's not "processing" the way we think of it today, it served the same purpose. Many of the plug-ins sold by people like Waves, UA, and of course the console stuff in SONAR are intended to recreate this processing.

Here's a good example of using of printing with effects. I've recorded a lot of classical guitar projects, and one guitarist in particular had a gorgeous guitar that projected like crazy on stage because of a massive resonance (around 200 Hz IIRC). Setting record levels to accommodate the peaks at this frequency meant everything else had to be quite low, and of course, tape hiss was an issue with classical guitar. So I cut around 200 Hz on the way in. If I needed to restore that a bit on playback - although I seldom did, because I wanted the sound to be more balanced - I could increase the bass in just that range. If I had cut it while mixing, the hiss from the low level signals would have still been present...actually, more prominent because then I would have needed to raise the overall guitar level somewhat. 

Let's also don't forget that in most cases (I'd venture almost surely on the stems the OP describes) the musicians themselves are supremely talented. The "A List" studio guys have a knack for playing just the right parts in just the right way, that's why they're A Listers.
I recall watching a video of Fagen and Becker talking about recording "Kid Charlemagne". They tried a bunch of different guitarists for the solo and none of them could get just the right feel. Then they called Larry Carlton and he came in and nailed it in one take. Imagine that; an iconic, killer solo completely off the cuff, one take and see you later, put the check in the mail. Maybe a once in a lifetime for most of us, business as usual for those guys.
The same applies to the engineers that record those performances. Usually they're the tops in the field and have access to great equipment and spaces that most of us can only dream about.
This is absolutely not a knock on anybody around here, we have some amazing talent on these pages. But let's face it, those guys are in another league.

i think we all agree that great performers performing a great arrangement leads to an easy and probably great mix, but which plugins should i be using!? 

Anderton
As to the guitar example, "back in the day" the intention in the studio was often to capture the live sound of a band. A lot of those musicians had effects they used onstage, they brought them into the studio, and recorded the resulting sounds. So while technically speaking processing may not have been added from the studio's backline, the tracks were still being printed with processing. Also with something like drums, the room itself added processing, and engineers took advantage of that "effect" with mic placement. 
 
So the bottom line was that a lot of the work required to get the sounds people wanted occurred before the signal hit the tape, and while the signal hit the tape. Although it's not "processing" the way we think of it today, it served the same purpose. Many of the plug-ins sold by people like Waves, UA, and of course the console stuff in SONAR are intended to recreate this processing.

+1
 
I really couldn't have said this better myself Craig and I know you lived through it that's why you understand it very well.
 
I know that people have a hard time believing it But trust me, things have changed a lot since then.
A good example is Aerosmith's "Toys in the attic". Almost everything you hear was printed right to tape. As was the follow up "Rocks". which was recorded with the rolling stones mobile in Waltham, Ma.
Toys was mixed and recorded at the Record plant by Jack Douglas (and Jay Messina) Jay still mixes now. you can find him online and his rates are reasonable.
If you really want to know what it was like recording and mixing a classic album back in the day, this is a great inside look.
http://www.aerosmithtemple.com/latest/how-the-bad-boys-from-boston-recorded-their-finest-albums
 
What reverbs did you have at the time of recording Toys in the attic ?
"They had some really cool EMT plates, and a spring reverb. Most of the reverb you hear—like on “Sweet Emotion”—was the EMTs. If we had a cool reverb sound going, we just printed it to tape."

Incredible insight on here from everyone.  I'm just sitting back and getting educated.
 
Most of the music circles around here (Nashville) are looking for the guys -- ahem -- one guy -- who can do it all...write, sing, produce, mix...the works.  From a computer.  So...that's what I'm trying to get better at.  I'll never be a Larry Carlton or Quincy Jones anyway, so it works out...lol
 
This video of Charlie Puth is pretty unreal...say what you will about pop music, but this kid is a prodigy.
 
https://www.youtube.com/watch?v=IU8BEMi8UyM

This thread should be a sticky.
One of the best threads I have read on these forums:

Great contribution and reference.

"One guy who can do it all" is the exception rather than the norm, and "from a computer" sort of pushes the editing tools. The DAW/app industry banks on people striving for the one guy aspect. Overdubbing and live performance are different beasts, although there are some bands that were formed to perform one person's work.