I think it's also important to recognize how much survivorship bias there is in these stories, too, though.
There are plenty of people who have spent similar amounts of time working on games or projects with nothing much to show for it. The experience might be worth something to someone, though, in terms of resume building, etc.
Games are really easy to test viability for, though. The game has perfect information about itself, rarely connects in a complex way to anything outside the game, and it's easy to replicate the user's context just by sitting in a dark room on your own. The test is simple - is it fun?
By contrast, most apps need to connect to people, software and things outside themselves - they're part of complex systems, and it's really hard to know if they work well without putting them into use and seeing how they perform - and that means selling and talking to users, not something you can do in isolation.
It can take a long time to get enough of the game together to test fun vs not fun. Racing games for example can be play tested before the art is finished, but they are heavily dependent on subtle interactions between the tracks, cars, physics engine, and controls.
Yes this is the confusing thing about creative endeavours. Lean start up says to learn quickly and pivot yet so many artists work on the movies, games, and books for years and years.
I guess we need to realize that most of us are not James Cameron :)
You wider point is reasonable though as many people waste a lot of time building something that never gets traction.