Another factor is in the USA, at the undergraduate level, there is a strong incentive for "professors" (more assistant and adjunct than tenured) to grade in a manner which is totally rationalizable and consistent between students. Administrators call students "customers" who are buying a degree; if a student doesn't like their grade they will complain about it, and if the teacher hasn't explicitly formalized their grading process they will be in hot water. So for example asking different probing questions of different students might lead to accusations of arbitrary and capricious grading. It encourages checklist-style rubrics for grading. Obviously one could formalize a decision tree for more loosely structured oral exams but that's significant effort.