[QuercusMax]

Sounds very similar to the problem of accurately assessing performance of software developers working as a team.

Sure, you have people who write a ton of code and implement a lot of features, and fix a lot of bugs, who clearly are contributing a lot. But you also have people who, though other means (code reviews, refactoring and other code-health work, etc) ensure that a project is maintainable and sustainable.

How do you measure the value of 100 bugs that never made it to production because of high quality code-reviews? Or those 5 high-value features which were a snap to implement because somebody took the time to clean up all the cruft from Mr Rockstar Bro who made a gigantic mess?