Space

A minute long test sounds awfully short for something that will cost millions if it fails.

Testing in real life when failure is so expensive is less than ideal, any controlled closed environment is better if it means you avoid failure. The very next paragraph from the one I quoted mentions how another rocket spent weeks in testing this specific matter and was delayed because they found the issue.

I'm paraphrasing what i read some NASA dude said about spaceX, but basically, if they failed as much as spaceX did they would be out of their jobs yesterday. Also, you know who takes the cost of these failures right? It's the US government through all the expensive spaceX contracts and tax breaks they sign. Each blown rocket makes the contract renewal more expensive.