I have another conclusion. Here, for example:
"they would deviate from set procedures because the procedures can be “long and winding.” He said “sometimes you just know that you can adjust a valve slightly and that would solve the problem. Following the procedures might take you down a long and winding path.” You might think that's no big deal, just an engineer using his engineering skills; if he was caught doing this kind of action on an American submarine it would cost him his job and possibly his freedom. If you work on the Strategic Weapon System you must follow the procedures, mistakes can be catastrophic."
What I read is just that he doesn't understand the issues. The engineers make something and make the consoles sounding "alarm" for many various reasons. They also make the documents of the "standard procedures" to follow.
Now some alarms are in use observed to happen too often. That's a clear problem in how the thing was designed. The cleanest response would be to redesign the devices, or at least the documents of the standard procedures to include the "shortcut checks."
But in his specific case it's just the people on duty that passed their knowledge from one to another. Maybe they even tried to initiate change in the documents or even asked for the redesign and failed.
So they are not wrong for doing what they are doing, it's just that the system doesn't care enough for the feedback from "the trenches."
Which happens everywhere.
In that light, all non-engineers or people not being "in the trenches" with the technical equipment really have too idealized understanding that the things "just work." No they don't, everything fails all the time.
Which is why we really should press the politicians to reduce the number of nuclear weapons. Sooner or later there will be some very nasty failure, an unintended accident, and it's a big difference if that failure doesn't destroy the whole civilization at once.