The design-build-test cycle necessary for physical products where the build and test phases can take days, weeks, or years was being used for software where build/test can take seconds or at most hours -- faster design cycles are beneficial and lots of folks used to engineering things got stuck with really long cycles. You don't need a management framework to accomplish those faster cycles.
The only thing needed is training on proper usage of the tooling. But that can come about organically when companies accidently hire a few competent engineers.
Isn't that quite well in accordance of Agile Manifesto's "Responding to change over following a plan"?
Sprints and backlogs and grooming, all those sound they're straight from the Scrum Book. And while I understand that for many Scrum == Agile, actually it's not.
Note that I'm not saying anything about what's the best way to do software. Only that "responding to change" sounds very much agile.
I see Scrum a lot like I see Six Sigma. They take some of the processes you see from a successful team and they turn them into gospel for managing all teams. Sometimes they try to apply these processes to the rest of the company in places where they don't make sense. They create a bunch of certification levels so that they can claim you can't just learn it from a book. They make a bunch of money but leave their clients in a strange place.
The waterfall paper criticised a process that did exist and still does exist in some companies - it didn’t make it up.