Rocketry is kinda different. Testing to failure can be very useful, and if you have the resources to throw at it repeatedly, can let you iterate much faster.
You can only pick two:
speed
quality
cost
NASA usually picks quality… and nothing else. SpaceX picked speed and quality.