Talking off the top of my head and theoretically (and also thinking that I read something similar on these boards before), if you had conditions that had a very coloured light that you wanted to 'balance out', or if you wanted to introduce a colour-cast, wouldn't you be better filtering it at capture rather than pushing one of the colour channels to extremes either in camera or on computer?
I can't think of any real-life situation where this might be the case, but I'm throwing it up for theoretical discussion in the hope that someone with more knowledge/experience can enlighten me.