But, what utility could there be for the optimization you describe?
None, but then there is no point in sorting in a void context either, but I've been bitten enough times by that when benchmarking (usually publically!), that extreme differences make me suspicious. I was just suprised by the magniitude of the difference you were showing.
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
Lingua non convalesco, consenesco et abolesco.
Rule 1 has a caveat! -- Who broke the cabal?