- Simracing: getting faster doesn't make you win races. Getting more consistent does.
- Squash: getting stronger and faster doesn't make you win games, getting less exhausted does, because you make less mistakes.
These are surely not true when you're at a high level, but for most people it's the right focus.
The corresponding insight for building software is this: "Brilliant" ideas for your engineering are probably not the right place to invest energy. Instead you are likely to create more success by ensuring you have excellent testing, qualification, user feedback, monitoring, rollouts, stuff like that.
This also just comes back to investment+return again. If _you_ think you gave great ideas then that's fine. But if you can keep your race/game/project/startup alive and stable then you start to unlock extra "passive income" from your teammates having good ideas (and the latitude to execute them) and your competitors fumbling their opportunities.
So take your ego out of the equation and do the humble stuff first. Don't reward your "rockstar programmers", reward the person who set up the dashboards and the e2e tests and threw up together that spreadsheet of 'most common user complaints'.