If you want to perform a modelling of ISFET’s you should apply the right theory about their working principle (how and why is the selectivity achieved). I don’t know the SPICE model but I am wondering that it is based on an accepted electrochemical theory explaining the selective behavior, since even electrochemists are still arguing on the right theory (with or without including kinetic aspects, e.g., exchange current densities). Another aspect is the sensitivity of the semiconducting layers against the migration of alkali ions in the latter changing their properties drastically. We were not able to prevent this deterioration of the FET and stopped working in this direction. Without demonstrating that your model is valid in the real world and works in practical applications, the modelling is useless. Just statements without any proof is not very scientific.

In my humble opinion, it'll take at least 6 months from the time you enter the cleanroom to when you have your first working ISFET. This is assuming your cleanroom is of decent quality and you have access to good training. Depending on what you plan on doing, this may be a worthwhile exercise or a waste of time.

If you're only interested in modeling, then you may not need to fabricate the device. You can rely on device simulators such as TCAD Sentaurus, which will nicely capture the transistor action of the ISFETs. The ion-sensitive gate material is not something that Sentaurus can model natively, but you should be able to modify some existing models (e.g. traps) to get what you need. There is a wealth of experimental ISFET data out there for any gate material that you might want to model.