It would seem that programming the wheel to a circumference to 5.000 and mounting two magnets would solve the issue of the display accuracy, and remove one variable (and thus one more potential error) in measuring.
What are your thoughts on this procedure?