I don’t doubt that at all. “Good” is relative to something but what that thing is often makes for a poor quality of measure with few samples compared.

The are objective ways to measure competence in software but nobody wants to use them. Most people don’t want to discover just how good they aren’t.