GIGO
I have worked with the modelers in NCAR, although it is now a decade old.
To give them credit, they state they are not attempting to create "routine forecasts".
What they are attempting to do is to see how dynamics of potential situations may work.
My best judgement is this is too far from reality, based on too little real data, and too little actual information about the potential interactions, to actually be useful.
Keeps the modelers busy and well paid, though.