Imagine a business that sells deliverable medical testing kits. This business is the best at what they do; they sell more kits than any other organization, they cost less, they're delivered faster, and are superior in any way, shape, KPI, or form imaginable to traditional alternatives. The only issue being that each kit is in fact a bloodletting kit, and their intended use is to inform you which of your four humours is out of balance.
I'm of the opinion that e-proctoring, whatever its privacy concerns, is fundamentally an exercise in "doing the wrong thing faster". Our exams are a poor reflection of student ability. Our exams have no bearing on actual proficiency in the subject matter. Our exams are ineffectual at catching cheaters. Leaving aside any discussion of privacy, e-proctoring (and traditional proctoring) fails to accomplish its fundamental goals.
To automate these exams, to make them ever-more scalable and easier to distribute, is not a victory. We've merely perpetuated a flawed system.
If we want a meaningful assessment of student ability then we need to use better methods. Dethrone exams from our curriculum. Leverage project-based assessments. Use oral exams where feasible. Replace the infinitely looping lecture halls with recorded videos and open-source textbooks. Use the recovered instructor time for something meaningful. This is not a Gordian knot. It can be solved with a little courage and a little pragmatism.