It sounds to me like a good example of "goal displacement". Someone has a goal and everybody agrees on measures, task, etc. And then the original goal is getting displaced with "getting the measure to X" or "delivery of task M on time, on budget". The goal itself gets lost along the way.
Hence, I think the problem here isn't impact mapping, but the goal got lost. Stop wasting effort here until someone can double check whether the goal is reached or not. Or define a new goal. But replacing a system for its own clone is not a good gool.
If you would own the company you are working in: would you like to see your money spent on your activity here? If you wouldn't, then don't do it. At least communicate it up - there is a small chance that someone up in the hierarchy can remember the goal and get you back on track.
(just my 2p)