While obviously better to get it in-camera, i am often white balancing in post.
Dealing with photography in Photoshop, when obvious whites/blacks cannot be found for balancing, i :
1.duplicate the photo (command+J)
2.inverse the photo duplicate on the upper layer (command+I)
3.change the composite mode to 'difference'
Now any area of the photo that is 50% grey will show as at least 99% black, so then i:
4.put a color sample marker on that spot
6.hide the top photo
7.open a curves adjustment layer
8.use the eyedropper from the curves layer that sets the grey point and click it on that spot.
sometimes i pick a few spots to find the best result.

in FCP for video i:
1.duplicate clip and move it above the original
2.put an 'invert' filter on it
3.change composite mode to 'difference'
4.zoom in on the blackest spot and toggle between the clips to pick the 'best' grey
5.apply 3way CC to bottom clip, disable top clip (or just pull it shorter so i can go forward/back one frame between clips) and click the mids eyedropper on that spot.

does anyone else do this, and are you getting good results? it seems a good way to get a decent color balance in tandem with using the scopes and eyes (which, with less than ideal monitoring is difficult!)

Why on earth would anybody do that? That makes no sense. If you want to white balance, use a white card or chart. If you can't, then I can't believe that jumping throug all the hoops you've outlined will do any better than just using your vectorscope.

i was trying to match three different cameras and this technique gave me a pretty good starting point to work from. it is really not so many hoops to jump through...

like i said, of course it is better to get it in camera, but this technique helps.

for white balance in post, i use the white and black eyedroppers in 3 way CC to click on white and black objects for a starting point. if white / black objects can't be found, then i use 50% grey with the above technique, also using eyes and scopes.
is this somehow wrong?