After reading the replies posted so far, which I greatly appreciate, I feel it would be best for me to make a clarification.  I probably misrepresented Graham Cochran’s teaching, so let me explain.
 
Part 1 of his Smart Start to Mixing tutorial is called, “Plugin Sweet Spots”.  He opens up a new mix from a client and presses play.  The Master Buss instantly goes into clipping.  The individual tracks are fine in and of themselves but the sum of all these tracks is way too hot.  He then inserts a trim plugin on each track (the equivalent of our Sonar Gain knob) and reduces each track by at least 10dB.  He presses play again, and now that Master Buss is under control.
 
However, there is a secondary reason for the reduction of each track.  He states that there is a sweet spot for using plugins which is an input signal around the -18dB level.
 
That brings me back to one of my original questions.  If the initial aim in working in the mixing environment is to work with tracks in the -18dB range, should that then be the target level in tracking audio?  Forgive me for implying that there was an actual “Sweet Spot” in recording a digital track.  As long as it’s not clipping, then digital is digital, or so it seems.  The analogy I gave in my original post was questioning an equivalent red zone on an analog VU meter as far as digital was concerned.  Maybe it’s best to target -18dB in tracking?  Hence, the idea of a “Sweet Spot”.  That way, any Gain knob usage would be minimal.  What do you think?
 
Thanks again.
 

Just to derail a bit, I can't understand why the meters in the prochannel console emulation aren't proper VU-meters, which would be very helpful in gain staging, as described in this thread. Posted an idea in the bakery here:
http://bakery.cakewalk.com/Idea/14457/VU-meters-in-the-Prochannels-console-emulation

The unique quality of digital is it is capable of perfectly linear amplification.
Analog gear simply can't achieve "perfect" linearity.
Every component in the signal chain is changing the signal.
Consequently every piece of analog gear has its own sonic "signature".
Plugins that "emulate" classic analog gear are trying to get that "signature" right.
So, do some plugins have a so called "sweet spot"?
I guess if the emulation is "spot on" they should.
There's an excellent page here that tries to explain dBu, dBV, etc.
It all comes down to "What are you hearing"?
T

RobWS
After reading the replies posted so far, which I greatly appreciate, I feel it would be best for me to make a clarification.  I probably misrepresented Graham Cochran’s teaching, so let me explain.
 
Part 1 of his Smart Start to Mixing tutorial is called, “Plugin Sweet Spots”.  He opens up a new mix from a client and presses play.  The Master Buss instantly goes into clipping.  The individual tracks are fine in and of themselves but the sum of all these tracks is way too hot.  He then inserts a trim plugin on each track (the equivalent of our Sonar Gain knob) and reduces each track by at least 10dB.  He presses play again, and now that Master Buss is under control.
 
However, there is a secondary reason for the reduction of each track.  He states that there is a sweet spot for using plugins which is an input signal around the -18dB level.
 
That brings me back to one of my original questions.  If the initial aim in working in the mixing environment is to work with tracks in the -18dB range, should that then be the target level in tracking audio?  Forgive me for implying that there was an actual “Sweet Spot” in recording a digital track.  As long as it’s not clipping, then digital is digital, or so it seems.  The analogy I gave in my original post was questioning an equivalent red zone on an analog VU meter as far as digital was concerned.  Maybe it’s best to target -18dB in tracking?  Hence, the idea of a “Sweet Spot”.  That way, any Gain knob usage would be minimal.  What do you think?
 
Thanks again.
 

I can relate. A while back, I was having an issue with signals going into plug-ins. For instance, plugs like Waves CLA Vocals.......there's a light on the input signal on the plug-in UI. This light changes from off, to green, to yellow, to red, depending on the signal being received from the track. Green is fine, yellow is best, red is too hot. However, if your input signal is too weak going into the plug-in, it won't get to the proper level for the plug-in to do it's "magic" as designed. So my vocal signal was not even registering in the green, even with the input cranked. Therefore, the track was not recorded at a high enough level to even register on the plug-in, and therefore, the plug wasn't even doing anything. So, I re-recorded the track, peaking around -12dB, put on the plug-in, and it was in the green and yellow, and the plug was doing it's thing.
 
Same thing for other plugs like Amplitube. You gotta have a strong signal going in to get the best possible sound out of the plug-in. So, there is kind of a "sweet spot" or a minimum signal level required to get the plug-in to actually do what it was designed to do.
 
It takes a lot of trial and error to reconcile all the different variables of input levels (interface, DAW, plug-in).......but usually if you get the input level correct on your interface, (I use -12db), that will drive your plug-ins to provide the best sound.
 
~Jeff

DeeringAmps
The "sweet spot " is analog, not digital. The old notion of recording as "hot as possible without out going over" was always a bad idea in my experience. Full scale digital 0dbfs is a hot signal. Most prosumer gear is "edgy" at best at those levels. -18 or so seems to be 0db, analog meter right on the red. With 24 bit you can run well below that and still keep your s/n at good levels. -18 was still the right level at 16 bit, but the "pros" were too dense to figure that out. Well designed and built analog gear can have 24db of "headroom " but not most. At least that has been my experience. Digital is not cold or harsh; it's just accurate. Garbage in garbage out.
Keep your nominal levels around -16 or -18 and the occasional peaks will take care of themselves.
T

 
This +1 ( db)  
I love where you say "edgy" and this is exactly how I would describe what I hear too when pushing the converters of my Tascam us1641. Some things you can get away with low end gear but signal paths into a recording are still the same as they always where except that last stage where it hits the converter. This is where we have to be the most careful of as you say, the recording becomes edgy, even if it doesn't clip.
 
 
Here's a example of this relationship between our hardware levels and Sonars levels. 
 
 I was tracking a vocal last night. I was using my regular set up, a Beta 58 via Joe Meek 3Q into the line ins of my interface.  The Joe meek has 2 outputs , one goes to the rear Ch 11 line input of the us1641 the other to a Mackie Mix 8 I use for monitoring. My interfaces 3/4 outputs feed a stereo channel of the mix 8 so easy to set any headphone balance. The mike sounded excellent in the headphones. 
The 3Q has good metering both for input and output. So my levels are easily set right there. no guesswork.  The 3Q's meters showed well within a safe level but Sonars meter was clipping? 
This is why we need to only trust Sonars meters. I quickly realised the little toggle for +4/-10 on the backs the 3 Q and the us1641 where you can't see them,, where not matched up. I set them both for +4 Now the Joe Meek and Sonar agree on levels. I had forgotten all about those little toggles. 
 
 

I was thinking about plugins that are emulations of analog gear. They may indeed have a "sweet spot". However I have not paid much attention to that idea. I don't recall any developer noting that in their literature. Maybe I missed it. 
 
 

I'll make one final comment and then put this one to rest.
 
The gain knob has a +18 -18 dB range.  It makes me wonder why Cakewalk settled on 18dB?  
 

there might be plugins that work best around -18db.
Personally would try to record the best sound/tone even if thats under or over -18db, and then change the gain inside Sonar for any such plugin that needs -18db.
That should get better results, just record what sounds best in my opinion.
 
 

RobWS
I'll make one final comment and then put this one to rest.
 
The gain knob has a +18 -18 dB range.  It makes me wonder why Cakewalk settled on 18dB?  
 

what I don't understand is why Graham inserts a trim plugin on each track ?
Cant he simply use (as cakewalk has) the gain knob ?
this is what feeds the plug ins signal. it is the first thing in the chain. So regardless whether or not your recording level is at this "Sweet spot" is basically irrelevant (unless of course its ridiculously over or under). 
Because you can always control the amount of signal that goes to the plug in by using the gain control (for those in the old analog days, its basically a trim pot).