The argument is that giving weight to the grades earned throughout the course may under-value students who made significant improvements. For example, take two

The argument is that giving weight to the grades earned throughout the course may under-value students who made significant improvements. For example, take two students: Student A comes in barely meeting the prerequisites and struggles with low grades on initial tests and projects. They come to office hours, put in substantial effort, and eventually ace the final projects and tests. Student B comes in with substantial background experience and gets decent but not stellar grades on all tests and projects. Depending on how much weight is given to the final tests/projects student B may still end up with a higher grade despite the fact that they probably do not have as strong an understanding of the material as student A. I've had both kinds of students the handful of times I taught graduate-level seminars and it's one reason I weighted the final projects more heavily than initial projects. (Though I found grades on initial projects were still useful for other reasons.)