The fact that there is an entire phenomenon known as "the unity look" points very strongly at some kind of exponential decay function in skill acquisition over time.

I'm having to fight for my life to avoid relying on the real-time lights in a relatively simple Unity project. I've written custom code to move chunks of scenes to origin to fix the light mapper range and sampling issues. This is the kind of thing where 99.99% of indie devs would throw their hands up and switch everything to real-time mode (resulting in "the look"). And, they might be right to do so. It could be the correct decision, but I worry a lot these days with how tight the market is. Steam sees a ton of new games every day. Having distinct visuals can make a really big difference. Selling <100 copies of something you spent >1000 hours on is probably the most catastrophic outcome for morale. Play testing and good mechanics can sadly only make you so much money if everything else is lacking. In 2015 you could get away with it. Not in 2025. The cheap ugly games happen inside places like Roblox now. You'll never compete with that.