And if you don't think the airlines would love to lobby Congress about the regulatory backstop, well . . .
The main reason why airlines would like to take the least amount of fuel is because it immediately increases payload capacity and thus flight efficiency. This being a cut-throat market there is a serious incentive to cut it as fine as possible. So the regulations around this particular issue are incredibly strict: you have to have a certain amount of fuel left upon landing, you have to write up truthfully how much you still had left and you will be investigated without fail if you cut into the reserve. The good thing about unions here is that they help to make sure that pilots know they are safe reporting truthfully because the airlines can not retaliate if they would pressure the pilot to not report an incident (which all pilots would normally definitely do). So they're a factor, but it is the regulator that writes the rules here and they are super strict about this.
And that's immediately why the calculation of the estimate becomes so important: you now have 30 minutes (or 45, depending) of deadweight + the deadweight for two alternates and an x amount of time in a holding pattern, plus up to three go-arounds. That really adds up, so you have to do your best to get the calculation as close as possible to what it will be in practice without ever cutting into that reserve.
It took me the better part of a year and massive amount of learning to write a small amount of code + associated tests to pass certification. It also taught me more about software engineering (as opposed to development) than anything I did up to that point in time and it made me very wary about our normal software development practices.
Like, I'm definitively not an engineer, nor does my day job really involve engineering, yet my title contains Engineer! I'm a proud CRUD monkey and designer.
I have done engineering work previously when developing hardware, and it's really a different mindset (even in an agile & fast-moving engineering org). Safety, cost, reliability, multidisciplinary integration, etc. just don't really come up in a lot in web and app development (which is a wonderful thing, really—I love it!)
> I wonder if there would be any value in teaching an "engineering when lives are on the line" or "war stories from accident investigations" classes to new engineers.
There would be immense value in that. But who is going to pay for it? It's a course that will essentially cause your crew to start producing software at 1/10th the rate they would otherwise do.
You'd think, but individual humans are very very bad at estimating risk, and in toxic group and work situations, humans will often take on increased personal risk rather than risk conflict. I.e., they will value group conformity over their own safety ... especially if their paycheck is involved. Fear of death is not nearly as powerful as robust regulation and unions.