You see this trap in tech companies as well as game studios. I call it "acting like Google before you're Google". And in fact virtually nobody becomes Google.
Google has been a money printing machine for 20+ years almost unparalleled in human history. That's allowed Google to write bespoke software for everything, which has been useful because almost nobody has Google's problems. It's also allowed Google to contribute a bunch to open source and engage in vanity projects. They can afford it.
Then you see the likes of Twitter a decade or more ago who dedicated possibly hundreds of engineers to make Cassandra work. That's doing Google shit. But they aren't Google. And eventually those chickens come home to roost.
Games are like any content business where the owners and leadership are trying to create a formula that can be repeated infinitely. Content business actually hate the creative people who make their content because creativity doesn't scale. This is why movie studios churn out sequels and superhero movies. It's a formula.
Games eventually fall out of favor as genres get stale and new genres get created. Minecraft is an almost unique exception to this. There's a reason it sold for ~$2 billion. It's still popular. It's crazy. But that kind of example is so incredibly rare you should assume it's never going to happen.
The hubris at Epic was that they could challenge the Apple and Google app store monopolies. They were wrong. And they wasted an extraordinary amount of time, money and opportunity chasing that. That was a strategic mistake, even though I agree with their philosophical position.
I'm reminded of id software. John Carmack was legendary for years. Wolfenstein was groundbreaking. So was Doom then Quake. But eventually (IMHO) id games ceased to be games but because tech demos to sell engine licenses before ultimately being acquired and swallowed.
I feel like Epic is the same with the Unreal Engine. Fortnite is a success while it's popular and people buy cosmetics but when that popularity wanes, they have a huge revenue problem.
> Then you see the likes of Twitter a decade or more ago who dedicated possibly hundreds of engineers to make Cassandra work. That's doing Google shit. But they aren't Google. And eventually those chickens come home to roost.
Isn't it the other way around? Using off-the-shelf solutions like Cassandra didn't work, so they had to resort to doing actual Google shit, a custom solution, to meet their needs.