A clarification. Vocal sync and all region fx for that matter are applied at the clip level which are pre effects bin. The exception is if you have clip fx, which will be bounced when you create the region.
Once you create the vocalsync region and have got it to the point you like it you should render it to get the time stretching in high quality, since otherwise its using lower quality realtime stretching.
Either way you are free to add effects and panning as you wish since it has no effect on the vocalsync processing itself.

Noel Borthwick [Cakewalk]
A clarification. Vocal sync and all region fx for that matter are applied at the clip level which are pre effects bin.
 
Either way you are free to add effects and panning as you wish since it has no effect on the vocalsync processing itself.

And another clarification to the clarification...the comment about effects is true if the processing is applied solely within SONAR, but some people record vocals through hardware processing they particularly like (e.g., the TC-Helicon or DigiTech processors). VocalSync doesn't seem to like those, so the solution is kind of like re-amping...monitor through the effect while you send the dry signal to SONAR. Then you can use the external insert on mixdown to get the sound you want.
 
Another problem can occur when someone sweetens up the voice with processing, bounces the clip to itself or applies DSP to make it permanent, and then decides prior to mixing to add VocalSync. Doubled and ADT voices are particularly hard to deal with. So I think it's always best to work with just raw vocals if you plan to use VocalSync, and apply any processing afterward.
 
The +1 was to Swamptooth's point about not using effects when working with VocalSync. If you have something like reverb happening, that will obscure the "real" decay so if you're trying to align the ends of phrases, the reverb can throw you off.
 
I guess the bottom line is I feel it's safest with VocalSync to a) record any vocals without processing, and b) apply VocalSync as early as possible in the recording process.

Thank you for the great points and clarifications. You are amazing. One general question about vocals, if you don't mind: when you doubling the vocals, do you pan them L+R in the mix or center? I kind of like panned L+R but it sounds like 2 people sing still, I need one. lol. The purpose of doubling is to make it stronger, right, so pan center? Thank you.

Not sure if you're addressing Noel or me, but here are my opinions. Of course they are totally subjective and are neither "right" nor "wrong."
 
First you have to decide on the intended impact of the vocals. I did a disturbing song about a returning veteran where everything was happening inside his head, so I wanted the vocals to be more diffuse. But with a pop song, you probably want strong, upfront vocals. With trance, the female vocals are often very diffuse so they'll blend in with the track more instead of stand out...they become more like another instrument. 
 
That said...I feel the strongest vocal will always be a good, solid, single voice. Doubling can dilute the impact among two vocals. However, you can augment the single voice to keep the strong qualities of the single voice with the more complete sound of doubling.
 
The VX-64's creates electronic doubling that is ideal for augmenting because you can place it in the background with the Presence control while the Stereo control adjusts the width. I prefer to widen it a little bit from the center so the voice is still centered, but bigger:
 

 
A little delay is also good but I don't like the delay to interfere with the voice. So, if you look at the Filter Mode and Cutoff controls, it's basically a high pass filter so there's no delay energy in the fundamental vocal frequencies. The delay rides "above" the vocals and gives more definition.
 
I did a shiny pop song that really shows off using stereo imaging, delay, and doubling to differentiate voices. Click here and then go to 2:20. There's a single voice, augmented exactly as described above. It doesn't sound like a doubled voice, just a "big" single voice. At 2:34, there's quadrupling of the vocals (not electronic, actual singing) and the four voices are panned to cover a much wider part of the stereo spectrum. These sound full, but not "big." At 2:38, the solo voice comes back and the contrast with the panned voices is very clear--even though it's only one voice, it's a lot stronger than the four voices.
 
Hope this helps.

Cool.. I just saved your post as a Word doc for future reference.