A relatively new way of measuring player value in the NFL is WPA (and it's sister for Defensive players +WPA). Win probability Added is the measure of impact a player has on the game and effects his team's ability to win. This is a narrative stat, it gives us an idea of the impact of plays in the context of the game rather than in a vacuum.
While it is not predictive (it only shows what happened in the past). And is also still being tweaked and redefined as more information on game probabilities is assembled (allowing for better modeling). It can be useful here to see what kind of impact a player had on his team's season.
Lets take the example of Jonathan Vilma. Last year he was worth .35 +WPA in the regular season or less than half a win to the Saints. This is a measure of all of his positive WPA plays. It does not account for any negative plays that were his fault (or the fault of his teammates) and I've yet to see a stat that does. So at best, Vilma was worth a third of a win, unless his replacement is significantly worse the Saint's should not see a measurable impact in wins and losses. In Contrast to Vilma, Terrell Suggs' injury deprives the ravens of 2.63 WPA+ this coming year that they will have to replace, a far more important injury than Vilma's suspension.
Peyton Manning was another player mentioned in this discussion and out of curiosity I wanted to check to see if his absence was as big a deal as it was made out to be. In 2010 Peyton had a WPA of 2.89 (7th among quarterbacks). His replacements in 2011 combined for a WPA of -.56. A difference of 3.45. WPA suggests that Indianapolis would have struggled in 2011 regardless of Peyton's absence (This suggests they only would have won 5-6 games).
Let's be clear, this is the best effort attempt at determining the measurable impact of player loss. It is not definitive and it is not necessarily predictive. But it's a start towards developing some statistical analysis of player values.