I use version 5 (for streaming only right now), and from what I am told the 'local' DTP has to run in full screen. The problem is also that cropping the area you want on DTP takes time, while in VB you mark the top left and swipe to the lower right. With Skype group video (we use that for our show) as people drop and then maybe come back, the content of the modules (shots) changes because the screen changes, and you have to quickly recapture new shots. Fast with VB (less than 5 seconds each shot) vs WC (20 to 30 seconds if you are real good). We asked them to look into that :-)
Now, if we were running multiple Skype machines (on per guest) then yes, it will work.