Comment by fjiowf

Overall I agree, but it's important to realize how easily the impact metrics are gamed. Sometimes the hard parts are in the technical details of programming a specific problem, and doing large scale cross-org collaboration is far more well defined. Or even worse, cross-org collab can just be pure noise by people with no technical understanding defining goals and projects and promoting each other.

So much damage is done by the wrong people being constantly rewarded and promoted for creating noise by "leading" large pointless projects instead of doing the real work. Sometimes I've seen the new hire engineer solving the critical issues like problems in a data pipeline by collaborating across teams is actually having more impact that any of the senior engineers or management bsers.