If any software developer reading this was working on a similarly complex project — $100m budget, 200 strong team, etc. — and was asked a year or more in advance, to commit to an exact release deadline, a laundry list of features, a promise of zero major bugs, they would say it was impossible. They would say that the way to deliver such a project is through an agile and iterative process, without hard deadlines and with only gradual rollout of features to progressively larger numbers of users.
But AAA games need to be an "event". They require a huge marketing campaign with a specific release date. They need a completed, polished product that overnight goes from being played by a relatively minuscule number of users to everyone. This is impossible, but the economic realities of the games industry means companies try it anyway, again and again, with predictable results.
On the other hand, smaller, indie studios have created a much better model: early access. Players pay a reduced price for the chance to play and help shape incomplete versions of the game. The game is gradually expanded and refined, just as any game is nowadays post-official release, except this is an explicit part of the arrangement between players and developers. It's not a flawless approach, of course, but it's still a much saner way to create games. And hits like Hades prove that it doesn't mean you can still have a big, successful final release once you've got a finished product.