So it must be a proprietary chip? I found no info about what it could be, but could remove it from the board and follow traces and datasheets to try and find out. That takes a bit of dedication to do though, looking through a sea of datasheets. So normally people who do it have more experience than me, and can narrow it down from PCB layout and supply pins
I'm ambitious, and think there's still a chance for an aftermarket coil. It seems unlikely for minelab to make their own IC. So it could be an easily available IC that needs a little work on PCB design to make a working coil.
The patent for GM1000 talks about operating the coil in resonance, or possibly in a phase locked loop. A bit of R&D is probably needed to make coils of a different size. The passive components could be optimized for the standard coil design.
As an experiment, I could adapt a 12" detech coil to see how it behaves with the GM05 IC. If that works ok, then replicating the PCB would be the main challenge for aftermarketting, instead of needing more thorough R&D or reverse engineering. The IC has 5 input wires to the coil, Tx, Rx and shield. That'd be a cool experiment. I'd want aftermarket for the sake of getting deeper targets
Edit: I tried the 12" coil. It was detected and sensing targets, but for some reason couldn't get any deeper than the stock GM10 coil. I'm guessing the IC or the circuit on it is tuned for the GM05 coil it came from