Ron, I think because the coil to pole relationship is fixed (if the coil is tight) the detector can be tuned to ignore the pole. Similar to the requirement to wrap the wire in a tight stationary relationship to the coil. Just as a loose cable flopping around can cause falsing a moving pole would create problems (falsing).
The sensing of a target takes place because the magnetic lines of force change between the tuned-no- target condition and the target-present condition.
If no change is taking or has taken place from the initial tuned condition, no target present is assumbed. Any change represents a potential target, and further analysis can be done to determine other factors such as depth, metal type, distance to coil center, etc..
With an additional winding in the coil, some triangulation to the target can be calculated and an appoximate size may be determined.
At least I think this is how things work.
Anyone with a better description correct me at will.