I caught a hint of doubt on the "non-commital culture" from GP and I guess I agree with that. If they are specific models of behavior, shouldn't the model come

I caught a hint of doubt on the "non-commital culture" from GP and I guess I agree with that. If they are specific models of behavior, shouldn't the model come with validating observations? If the new model is very encompassing so that good enough observations are hard to come up with or their specificity would detract from the broad view, It might be excusable. Otherwise, I would think that the paper has to clearly show it's raison d'etre to it's stakeholders, most likely interested industry -- with observations. And why not, sometimes predictions are due, specially when they contradict the status quo. The predictions don't need the pretense to be certainly true, they just need to be consistent (or follow directly) from the model. I would expect predictions, just like a new physical theory or observation may accompany bold predictions following from it's premises.