It is certainly a a good idea to try to measure the performance of teachers relative to all other teachers, and the easiest way to measure this is through exam scores, so I understand why the approach is used, I just think it's somewhat counterproductive in that you're not measuring their ability to teach the kids so much as their ability to train them. But I'm not sure how you'd obtain quantitative data to determine teacher ranking other than this way, so I don't know what the alternative is.
I'd also argue that - again, specifically in my country - teaching at university is all wrong. A generation ago it was free for students to attend, they just needed high enough exam scores to meet the entry requirements, and at university they were taught for the purpose of learning. Nowadays that's all changed, there are now huge fees involved, kids have to pay a lot of money to attend university, the number of university places open has increased dramatically, so now practically anyone can attend, regardless of exam scores, so long as they're willing to either stump up the cash or else get saddled with debt. And because such high fees are involved, it's not really about learning for the sake of learning, it's primarily a transaction, paying cash for qualifications.
As you note that almost anyone can attend University, there are many University degrees that are worthless, or even less than worthless, such as gender studies.
In reference to your concerns about standardized tests, ideally, these tests would measure how well a student can apply what they were supposed to have learned. For example, instead of reciting facts, the student would need to apply concepts they learned to determine the answer to a question.