Being a scientist and looking at the code that is written daily (and aware of the bugs that are found), I certainly agree that there is horribly written code go

Being a scientist and looking at the code that is written daily (and aware of the bugs that are found), I certainly agree that there is horribly written code going around. Most scientists aren't encouraged to learn about coding, it's something you do to get the job done. But like you say, that doesn't mean it produces incorrect output. Hopefully there are enough cross checks that mistakes are caught before publication.