I have read the documentation, but I am none the wiser under what circumstances I might wish to use it.
 
What advantages does it bestow? Better sound?
 
Would this be an alternate solution to the "distortion" heard on synth plugs, as discussed
with Craig Anderton some months back and now addressed with the 96k sampling feature?

http://forum.cakewalk.com/FindPost/3328140

So if i understand Dan Cale correctly, it's the same as recording 16-bit vs. 24-bit.
And because 32-bit is already plenty the additional headroom is negligible in practice.
 
Thanks, scook, as always.

Your first comment indicates you do not understand correctly at all. Also makes your second statement incorrect.

I am open to correction.