My argument is that you can't tell the worth of the player if you completely divorce them from winning. You could do that if instead of 162 units of 9 innings with a defined positive or negative outcome for the context the player belongs to holding the context constant you simply had rolling totals for 1458 innings of play with no definitive positive or negative outcome at the end randomizing player contexts. But that's not how the game is played.
Again, the purpose of the game is to win. That's what fans come to see. That's why players compete. And again, I'm not against the development of statistics to better measure player performance, I just think that things have swayed too far to the side where the stats and what is lauded have little correlation to actual winning - case in point, Felix Hernandez winning the Cy Young. To some that was a victory, to me it was a disgrace.
To tie it back to the article I posted, I think that coach is a genius. He treats the onside kick as a faceoff with a 50/50 chance of getting the ball. Heck, my guess is it's even better than that for him because he's so far ahead of the curve. You have to wonder though, why haven't more coaches looked at kickoffs this way? It seems to me there's a fair amount of strategy involved that could make winning the onside better than 50% for the kicking team. Even if you disagree with his philosophy on punting (and I don't) why aren't more people buying in on the kickoff strategy?