Yep. I am working at such a place. In the financial industry. A single “migration project” (“from banking system A to our system B”) costs more than Microsoft will ever make, even when selling 5 mio copies of FS2020 at an average price of, say, 100$. For a single bank (that is not money we earn, just to make sure here - that’s what the entire project costs, with all consulting companies etc. involved ;))
In such a setup you can afford “4 eye checks” - and tons of integration and migration path tests.
And yet: “mistakes do happen”. It is simply naive to think that the person looking at the code will spot every tiny little detail such as a wrong “flaps uplift multiplier”, given that this other person is usually also another developer with some tight “time budget”. 4 eye checks do make sense - and my colleagues have spotted many “WTF! copy & paste mistakes” I made. But detecting “logical flaws” is very hard if you did not write the code yourself (especially if the “code domain” is not familiar to you! Many companies tout about “team work” and bla bla, but in many cases you have just one specialist working on a topic - hiring more developers, sorry, no budget).
That is something only projects can afford where people’s health and life (think NASA, medical instruments etc.) are at risk, and where the code basis is “small” (relatively speaking) and especially “the purpose of the software very well defined (and limited as well)”.
Here on the other hand we are talking about a game! Microsoft has to make money, with every copy they sell. And again, we are talking about a 150$ product (yes, in a high volume market: 2+ mio sold copies is impressive, especially for a “niche product” such as a flight simulator). And I am pretty sure they do have automated tests in place.
But certainly not for every single flight system of every aircraft.
There is a particular wrong image in your head about developers being lazy here: for the most of the time developers do what they can, with the resources (time) they are given. Personally I do not know a single developer who wouldn’t want to “stabilise the code first!” (yeah sure, there are average, good and top players out there - but everyone makes mistakes!) - instead of throwing yet another redesign of the “needs to look webish!” UI at the product.
Managers, project leaders and stakeholders on the other hand…