You and tieTYT have convinced me. :-) But when I worked at a dev shop around 2000-2005 we did have our own app server (begun in the 90s), with its own templating language (two actually), its own "NoSQL" database, and its own scripting language (all operators and function calls in Reverse Polish Notation). And it was written in C. It was pretty fun to work on it, actually. But there's no way I'd build something like that today. And I think my desire to make games is stronger than my desire to make a game engine.
I enjoy inventing things and completing challenges. I think this trait is common among engineers. Unfortunately it doesn't make much practical sense in a lot of situations. I have entered Ludum Dare multiple times and each time I failed to finish a game because I end up spending the whole weekend writing a cool particle system or whatever. When I'm left to my own devices without planning, who knows what kind of thing I'll end up making.
The decision to invent should be made intentionally, rather than impulsively. If your goal is to have fun, then have at it. But if your goal is to launch a product, it's not prudent.
