The thing you wish is counterproductive. The values for zdepth are max distance and min distance, in the middle it scales linearly. If you would to bind the target to one of the values, all values beyond the target would get clamped and the result would be wrong in post.
What you wanna do here is extract XYZ for the camera target.
What I did in the past is the following: MaxToAe the pos data of target, in after effects use the toComp expression to get 2D screen space data from 3D xyz. And parent the focus picker to the null with the 2D data.
What you get is post processed DOF, that focuses on the camera target, that you keyframed in Max.
You cannot change zdpeth itself in a meaningful way. That data is not ment for that, instead try to get the data of the focus itself.
