If you want to go all retro on us, you may want to give this one (ADT) a try. ;)

http://www.vacuumsound.de/plugins.html

+1


X1 producer has the VX64 vocal channel that has a doubler

Use 2 different types of mics and record both at the same time. Then use pannning, effects and/or nudge forward or backward as desired.
 
Or you can record another take. :-)

i have done this by stretching it to maximum (400%), rendering, then unstretching it back to 25% and rendering.

VX64 doubler is good for just adding a little extra to a chorus or something like that, but it does not sound like 2-3 people singing.  It just sounds like a stereo delay.

+1 to what konrad said. I use VX64 at times for voice overs when i need very subtle (and I stress the subtle) thickening. Anything more than a touch and it sounds like crap. Not sure how it would work for singing.

Hint for those new to Sonar: If you do use the VX64 (or other effect) you can open it up, enable WRITE on the track, and then click the ON/OFF button on the doubler while the track is running so it will go on and off automatically at the right places in the song.  Example: turn it on right before the chorus and off right after.

This works with other effects as well, of course.   I am embarrassed to admit that I had X1 for a while before I realized that automation would control anything besides faders and pan.  I learned the hard way by screwing around with the EQ while the track was running with WRITE enabled.

Not exactly on point, but here's a new software I just read about (if you've got an extra thousand bucks or so lying around):   http://www.synchroarts.com/

Of course none of the automated doublers sound as good as old-fashioned double tracking like the Beatles did on their early albums. 

Beatles sounded cool when they got it right.  A few of them are sloppy for timing, starting way off from each other, etc.   Still a testament to how well they sang on key that you don't hear much pitch variance.