Ian, I agree with your sentiment but find your system impractical. Sorry.
The constraints of the contest do not allow any entry to qualify as a masterpiece for one thing. That, coupled with the potential for less than average among the submissions, leaves a small range for judging.
I prefer the bell curve (having been subjected to its rigors in the dark and dreary past of law school hell). It's useful for comparing members of a set against each other without having to refer to outside measures.
The best of the bunch is a 10, the worst a 1, and the rest scored as compared to those. So, if all submissions are masterpieces, we can still pick among them. Ditto if they are all dogs (hypothetical example ONLY!)
And then ... Everyone who didn't enter gets a ZERO!
pootsie