I've never graded tests or papers, but I always assumed cheating would be obvious because if you go to chat with a student about the problem, they will not have anything to say.
The total time spent in creating a few good, easy to discuss questions, the answering of which would demonstrate understanding, and then reading them, thinking about them, and coming up with a grade, was probably actually less than the amount of time it took to create meaningful multiple choice questions that didn't have any ambiguity, and which weren't easy to intuit the answer even without understanding. Doubly so when we did away with "partial credit" answers, but instead made it so each question was 1 point (or otherwise all or nothing), you had, say, 5 of them, and what we really were looking for was a paragraph that showed understanding (rather than checking off boxes in a rubric of "mentioned A, B, and C"; short essay questions, if you will), an expectation we communicated to the students.
And that's aside from the actual project based grades, which were better still.
By the way, where I work, management pushes for stuff like proctoring, and more students, and "measurable" results, and so on. As a teacher, I don't care much about the whole grading show.
Then the students are not getting what they pay for.