Peer review and student evaluations together give a much better picture of a professor's ability, and recognize that professor's ability to truly challenge students, not merely parroting the general wisdom. Obviously, there are inviolable facts, and it's important to recognize them - but learning is more than understanding that which is generally accepted to be true.
For example, I'm forced to teach eigenvalues, but I don't get to teach about linear programming, even though I think linear programming is well worth knowing. That's because eigenvalues are on the syllabus, but not linear programming. This would not change if someone else, rather than me, designed my final exam.
All standardization does is give you a mechanism for comparing my students to yours.
Obviously you don't use a standardized test for research credits. But you do use it for already known material taught in class. In fact, many graduate schools do exactly this - the standardized exams are called "written qualifying exams."
If you make the class difficult, students will hate you. If you are easy, they like you. There are a few other tricks you can do to get better evaluations: sympathy, good looks, and jokes.
It doesn't matter if they learn more in your class than they would elsewhere, since we don't measure learning. If we really wanted to determine how effective the professor is, we would standardize exams and compare student performance across sections. We don't do this, probably because we don't really want to know the results.