Not really. A minor problem is that it fails to capture the fact that this process will also hold the qubit stable in the |1> state. More importantly, it fails to capture the limiting behavior of the success rate; it should move smoothly towards 100% as the measurement frequency increases but the ball-and-slope would instead jump discontinuously from not working to working as the hand went from being too-slow to fast-enough.

A better example might be... a series of polarizing filters? If you have a series of polarizers but skip all the way from vertical to horizontal with nothing in between, no light gets through. If you put a diagonal polarizer between the horizontal and vertical, some light gets through. If you have a big long series of very gradual steps from vertical polarizer to horizontal polarizer, almost all the light gets through. More frequent measurements causing the zeno effect is like to adding more gradations of polarizer direction causing the all-light-gets-through effect.