You address the issues quite well.

But continuing from a devil's advocate point of view...

This idea of yours will only help scientific based predictions and scientific based predictors. Most of the predictors I see actually find such scientific values reprehensible to their mode of thinking. In fact, they see "high science" as THE hinderance to any effort to successfully predict quakes. These types will not use your service due to the constraints placed upon them.

I guess what I'm saying is, your proposal preaches to the choir.

But then, if your goal is to provide a service to the scientific based predictor, and don't care about the fuzzy-wuzzy-touchy-feely predictor, then I think you will do well. In fact, I may even get back into the prediction game myself.

Brian