Older devs can more easily have 6 mo to a couple of years in savings. It's not fun to burn that down, but it's quite a different situation from having no savings and student loans.
Also, having a partner with whom you can talk about money, strategize about money, and share income / savings with is a huge advantage.
The interest on 130K of private student loans, to start.
I've often heard the "Unity look" criticized, but there is no Unity look. That's just a failure of art direction. I.e. any engine that supports shaders well doesn't have a particular look.
Because Unity and UDK are such time-savers, they're used in a lot of student projects. Often these have very low production values that bias people against these engines.
A good analogy is particle systems. The fundamental technology hasn't changed in 20 years. However the final look is extremely sensitive to iteration and art direction.
Having good artists is much more important than having good tech, much as it pains me to admit it.
As a seasoned developer with a family and mortgage, I don't see how you could go indie as a 21 year old. Your probability of achieving absolutely nothing approaches 100% (and the end result may be career damage that will carry through for the rest of your professional life).
This story really has nothing to do with the fact that the guy has a family or a mortgage, and exactly the same realities apply to a 21 year old with rent and car payments. So it's a little humorous seeing every 21 year old (speaking of many if not most of the comments on Reddit) trying to declare that his core issue, which is really wishful "it won't happen to me!" denial. If anything, the guy with the mortgage and family obligations should have the motivation to achieve what is necessary, whereas this guy seemed to make horrible choice after horrible choice.
The story here is simply that he had a gross inability to estimate (if, as his runway comes to an end, he has so little...wow). And despite the fact that you disparage 2d platformers and Unity, the reality is that Unity, for instance, does a dramatically better job as a basic engine (for essentially all sorts of games) than the overwhelming majority of developers will do on their own - if you aren't John Carmack, your probability of doing something better is incredibly unlikely.
This isn't about pride (oh, boastful "I'm too proud to use that..."), it's about raw stupidity. It's building your house by deciding that you'll first make your own drill because surely you can do it better than DeWalt. Starting your web app by first making a custom web browser. It is project planning and introspection gone perilously off track.
I agree with the rest of your post, but this is just silly.
Yet often people just...get a job again (such as in this story). Exactly as the non-established person does. It really isn't that much of a difference. Life goes on.
That just isn't compelling to me, but that is a personal thing. So I don't do it. But while I don't do it now because of my personal outlook (also considering I'm not interested in game dev because the industry eats people alive) I definitely wouldn't fathom risking my family for a pipe dream. And then build my own game engine.
My question is, if you are a game dev dissatisfied with being an office robot, why not get a few friends and make your own dev house? It would be "indie" but you could ask around publishers pitching ideas or doing a kickstarter with more people and a fixed funding goal, with a much better understanding that games take 500% more time than you expect minimum, and make something big. You have the credentials after a decade or more in the industry to attract attention, I'd hope.