If you want longevity and plan to create games for many decades to come, and if it is within your ability, it is wise to roll your own tech on the side. The reason is that you can never rely on big corporations, as they are never your friend in a capitalistic system. So you want to keep dependencies as low as feasable. For that, you should also employ an engine agnostic programming style. When Unreal is no more viable next, you can port the code to another engine without much hassle, including your own. This is especially true if the game is rather code heavy above anything else, having many custom systems in place. But the custom tools can still remain tied to Unity, as using those is not what you are going to be charged for (or could you?
)).
One major advantage of game-engines is cross-platform compatibility, of course. But that should be covered by upcoming SDL3 (except for consoles, I gues). It effectively upgrades SDL2 by a cross-platform vertex/fragment-shader. That's plenty of power for sophisticated visuals.
Personally, I have a bit of investment in Unity, to see how the games are assembled according to the "industry standard", and to evaluate if it is a good fit for a bigger project. I preferred Unity over Unreal because of the purity of its architecture; you start with a clean slate. And I would like to stay with Unity for now, but I can switch engines and check Unreal next, always knowing I can fallback to my own tech as the majority of systems are custom code anyway.
Having that said, I understand if you are an artist, or if you want to include cutting-edge features only that one engine X offers, then that's the risk you have to carefully evaluate.
Player
