Assuming your color analyzer is stable, I think that's your best approach. But then, no one has mentioned how much variation there is in paper, from batch to batch. I've never checked it, maybe Ralph has some ideas there.

Are you using first world manufactuered papers, or second or third world?