SpaceX has been intentionally stress testing these test vehicles to see how close they are to safety limits. None of what’s in use today will carry customers payloads to orbit, they’re all just experimental vehicles validating the approach & finding limits. Other companies use simulation more.

There are active discussions in the industry of what teaches you more, building experimental vehicles like this or more simulation and lab testing. The industry through early 2000s lost most first flights of expendable rockets and 2-4 more early on. SpaceX is moving faster and learning more…

It’s balance too here I’d argue. Between payload to orbit in as few flights as possible and moving as rapidly as possible at the expense of losing hardware No one really could build out a system and be as agile as SpaceX has with Starship in changes with purely simulations and stuff