It's easier to see the issue if you flip the graph, and add random points to it:
Consider A = a half-baked attempt, B = a decent effort, and C = striving for perfection.
The quality of whatever you're doing improves a fair bit when you go from A to B. But when you go from B to C, you're spending a lot of time and yet quality barely goes up. That time could be better spent elsewhere.
It's all about diminishing returns. And based on the existence of adages like Voltaire's "perfect is enemy of good", or my grandma's "the devil fiddled so much with his son's eye that pierced it", it seems to me that people know it on an intuitive level, and that it isn't just about businesses and production and programming. It's simply that people fail to recognise when it's good enough.