This paper does seem a little dodgy in its assumptions. We shouldn't dismiss military-based studies? Okay, (ish), but then the study doesn't measure happiness vs performance, it measures "how do you rate yourself" vs "the army's selection process for awarding medals".

Given medals at least partly exist to reward extremely optimistic behaviour at the expense of personal safety, shall we say, having an optimistic outlook could singularly apply to the field of warfare where individual optimism in the face of unlikely odds is a good thing.

This isn't sort of mindset we wouldn't want in our doctors or engineers or bankers, or almost anyone else.