Both approaches are valid- NASA is the “we have one shot to get this right” approach and SpaceX is the “launches for test data are cheap” approach. Both are valid
Will be interesting to know how many of those 1k changes are tested.
That's what the next launch is for.
So just like Microsoft: testing in production.
It's a test flight
The key to spacex's success is that they are not afraid of having rockets blow up during development. They launch, learn, iterate, launch again. I'm not sure how any sane person can argue with their results.