Any further discussion of declumping itself should be a different thread. But be sure to read these previous threads (2011, 2012, 2013), and these rulings (2011, 2013), before you even bother.
The point of me bringing it up is in the context of "sufficient randomization", and me pointing out...