"Responding to change" turned into a justification for never finishing features. Stanford researchers found scope creep institutionalized and relabeled as "sprint replanning," a leading cause of cost overruns