While I wasn't initially fond of his choice it sounds like appointing a bean counter was the right choice at this point in time. It sounds like Goldin had no friggin idea how to really accomplish his mantra of faster/better/cheaper. It seems as long as it looked that way he was happy without any attempt to analyse if it was really happening.

As a real world process optimizer I can vouch for measurement as a key part of any real effort to make things faster, better, and cheaper. You've got to track costs, quality, productivity, and evaluate performance of intermediate steps. When you do, the problems/opportunities if not the answers are usually obvious.

Regards,
Mark