Hi Richard,
All of the elements are there. PI, being a broadband system has the potential for very good discrimination. The received signal contains a large amount of information about the object in terms of the amplitude, width of the signal, decay time and the form of the decay curve. The Goldscan that I made some years ago was primarily a ground balancing PI, but it could notch out other targets and reject many iron objects. This was done by taking just two signal samples, but by taking more samples and having TX pulses of different amplitudes and widths, the system could be considerably refined.
Whether it is possible to discriminate, identify, reject, at the same ranges as it is possible to do basic detection, is doubtful. Generally speaking a higher signal to noise ratio is necessary to achieve these additional facilities. However, PI has the benefit of being able to transmit more power if needed, so I am confident that greater detection and discrimination ranges could be achieved than with VLF devices.
PI can also be used with balance coil technology and further signal samples taken during the transmitter on time. This is what the PPD1 used for Fe/Non-Fe discrimination. Combine this with multiple signal sampling in the off time and alternate TX widths and on times, and a very versatile detector could result.
Work is going on in all these different areas by engineers and researchers in several different countries, who are co-operating via the Internet. Shouldn