Spot on. You've described Peter Naur's "Programming as Theory Building" paper exactly.