It is very interesting the concept around the deadband you mention. I never heard it before. This and the fact i was working on the limits of minimum distance could have done the errors i had seeing. With the deadband in mind these errors are expected.
I think the movement with defocusing could be excluded because i defocused, then wait for some seconds, and then i shooted with a remote.
The "people" pictures I take are between 1 meter and 3 I calculate.