Because often it's not that hard for the employee to take credit for someone else's results or just fake them. It also creates a poor set of incentives. It causes problems like when programmers are given a bonus for fixing bugs, resulting in them adding bugs to their code and then fixing them.

Because results are tricky are measure. "Butts-in-seats" is much easier for lazy and clueless managers to keep track of, so that's what they do.

Yes. It's hard to measure results, because you usually want to know the counterfactuals, like: "what would the result have been if someone else had been given the assignment".

If sales grow 10x under your watch, was that because of your hard work and talent, or because the economy went through an upswing and a few lucrative customers stumbled in the door on their own accord?

If your colleagues responsible for the neighbouring regions aren't doing as well, would you have done any better in their place?

Very hard to know. But they can easily see if you're working overtime.

What about average or slightly-below-average performers who are putting in the coveted 12-hour workdays?

A raise and a big bonus!