Simply put, I find that my sweet spot is between -12 and -6 for a good, healthy signal. That leaves me enough headroom and if I need to turn down a track or two it won't disappear as fast. Also, I quit using Normalize years ago. I use the gain tool to adjust anemic tracks from time to time. I would rather adjust the gain in increments rather than peg it out with normalization. Most times if I use Normalize, it cranks up a lot of background noise if the track wasn't captured well.

I was just posting some related questions on the Software forum. Jeff, thanks for the detail here - I'm off to check out your other post!

Rarely do I ever bounce my V-vocal tracks before mastering, even if I have tons of vocals, so I can always go back and tweak a volume, the pitch or timing of a segment up or down.
 
I use the auto sensor button as I mentioned and sometimes (to be honest) I do get and over. I am not THAT perfect of a vocalist (hehe). But because of the 24 bits I am using I have found it can sometimes take an over and not create a horrible distortion if within a certain reasonable range. I make several takes of a vocal track anyway so I can comp and pick the best useable phrases. 
 
As for getting vocals to cut through a mix, I use the combination of three things. (on all vocals) A roughly 3db high pass shelf that goes down to about 1khz (+de-esser), I gather certain music tracks tracks (whatever works) into a sub bus and side-chain a compressor to nestle the vocals into the mix and I adjust the  bass total highpass cut shelf up to about 100hz. 100hz allows most warmth to remain in the vocal but prevents banding in the bass ranges.

That's why I won't use Sonars Normalization, You have little control. It is possibly the worst example of a normalizing tool I have ever used. Even Wavosaur has a better tool. 
If your hearing a hiss then either something was recorded badly, or the normalization is funky. There should not be any noise in a properly recorded track. And if there is then this points back to the old school reason for keeping your levels up and driving the pre amp properly as I'm saying here. 
 
Studio recordings are always fine because I'm sitting there in control, But with our live recordings I'm not always able to keep my eyes on the meters so I have keep the levels conservative. 
 
I tool copy to wave lab and run the analyzer first. If the level is 20 or higher average RMS I'll first normalize to -2 db, run the analyzer again and if it's around 14-17 it will be where I want it for mixing. I also will manually go through the track and knock down any obvious peaks.  This process will not degrade anything as some people are implying, if it does your using the wrong software.  
 
I too use the gain for tracks that are not loud enough and require pushing the fader past unity. But some tracks are just way to low from live recordings. 
 
PS- Why are you using V vocal for volume? I have never hear of using it that way. Because it is so buggy I only use it out of dire desperation on a tiny spot that needed pitch correction. 
First I'll steal from another place or take. 

Sonar has changed its normalization. It now has a setting for amount. This was done back in 8

I use V-vocal for my entire vocal tracks. It is buggy but by saving my work often (and multiple copies) I can get by.
 
I usually make two v-vocal waves per lead vocal track. One for the first verses and chorus and one for the last verse with possibly bridge and chorus.
 
Once my vocal track is (copied or resampled, not sure which) in V-vocal I can scrutinize every syllable for pitch, timing, formant and volume. I can (in real-time) sculpt the pitch throughout the vocal and stretch vocal passages so they have the right inflection. Thus I do not have to guess how many DB to raise each part with the gain option process effect. I can slowly nudge the many parameters in V-vocal of each passage 'til it sounds right. Then I will also apply a gain envelope to the whole track for global dynamics if needed.
 
I swear by V-vocal and I have spent thousands of hours editing in V-vocal. I live and breath V-vocal. It is BY FAR my number 1 plugin for Sonar.
 
V-vocal used to be a main selling point of why I would urge others to buy Sonar over other DAWS because it came already stock with Sonar.
 
BUT... not anymore. I guess Melodyne says it does the same thing but I think at least for now Melodyne is more buggy than V-vocal and (to me) has a clunkier interface.
 
In my opinion Fender ought to snatch up V-vocal at whatever cost and offer both options Melodyne and V-Vocal. What do I know I have only spent thousands of hours in V-vocal. 
 
I get the strangest thing that happens every time I edit in V-Vocal, I think of the X-files. I am not sure why but both the X-files and V-vocal are hardwired to the same node in my brain. Cool huh?    
 
I think it is from sunstroke from my computer monitor. :)