Facebook is one player in a large ecosystem of workers and companies, in which things like lunch and nice offices and health insurance are simply table stakes, and total compensation levels are broadly comparable. Until recently, Facebook did pay at the upper end of that spectrum, which was some combination of their stock doing well and people souring on social media as a force for good in the world. Certainly relative to peak social media excitement ca. 2006, working there is now considered going off to be a cog in a vaguely evil faceless machine; they couldn’t get away with lowballing people the way SpaceX or even Google can.
Hedonic adaptation is real. You compare yourself to your peer group, in which there’s always people living larger than you, stocks appreciated more than yours, bought their house earlier than you, higher earning spouse than yours (or any spouse at all if you’re single), generational wealth from China, etc. And homeownership in the Bay Area is such an insatiable black hole that this kind of money merely puts you in the running. You’ll never be, like, unable to repair a household appliance - which is better than many people! - but neither are you just waltzing through life milestones in the way people think when they see these figures. You’re mostly a pass-through vehicle from your company to local property owners.
Some companies are more top down and some companies are more entrepreneurial. Amazon is famous for assigning just the right amount of work to break you before your stock vests. Apple has rigid and precise opinions about what it wants built, with engineers discouraged from scratching their own itches. Facebook on the other hand is all about initiative, with engineers being almost like Wall Street traders: come up with ideas and implement them on your own, and in your performance review we’ll check the numbers to see whether you made us money or not. Like trading, you might have a good hypothesis that just didn’t pan out, or something else outside your control might have shifted, but that’s not going to save you. You have to be right. It’s stressful! But one thing that happens in places run this way is a pretty strong social norm against trying to stop anything before it happens. If someone wants to run an AB test, however stupid it seems, they get to run it, and you have to trust in the data (and data analysis) to reveal whether it was really a good idea or not.