Adding more won't help though. You can't truly gain anything through post-processing because the amount of information you have is finite, you can only make tradeoffs. In this case, you're trading spatial precision for more temporal precision.

Motion blur done properly wouldn't merely be post processing, it would be add the new information of the path (locations and dwell time) that the object took from location A to B. It's true there's a tradeoff here, you're obscuring the endpoint, but at sufficient Hz the error rounds to zero.

Mouse pointers actually are a good example of this. You might be seeing 4 of them, but at least they're all mouse pointers and not just a blurry white streak.

But if I wave a physical object around fast enough it looks like a blur, not the physical object. Making it look like the mouse-cursor at that speed is not an accurate representation of the implied physical model...