Your view of a lot of the methodology in science, and more so in engineering, is correct or nearly so, and I agree with it nearly so. BUT!!!! So far in saying

Your view of a lot of the methodology in science, and more so in engineering, is correct or nearly so, and I agree with it nearly so. BUT!!!! So far in saying what human sources of CO2 will do to the temperature of the earth, as in the graph at the link I gave, we're in deep, fuming, reeking, brown sticky stuff -- the predictions from nearly all the models were badly wrong. So, one approach to something better is to return to first principles of physics. History shows that that approach has been known to work. And IIRC by now the Navier-Stokes equations have been quite good for boat hulls and airplane wings. But as I outlined and you noticed, using first principles to say what the temperature of earth will be in 100 years is too difficult, absurdly too difficult, even if ignore turbulence around leaves, too darned difficult -- and it might even be unstable. So, that approach won't work, either. So, as you outlined, we need to find a way that does work. In such a case we will need to check it, test it, validate it, etc. And at this point, from the graphs, no one but 10 year old sweet, cute, darling, adorable but gullible and naive Virginia will believe any testing based on predicting old data and, instead, will want a prediction of the NEXT 20 years and the WAIT 20 years to take the work seriously. Soooo, right, we fall back to some back of the envelope approaches. Here's one such I've already explained here: (1) For CO2, for nearly everything in the record for the past 800,000 years (the geologic time intervals in the lecture seem to be different) shows that temperature went up without preceding CO2 concentrations going up and temperature went down without preceding CO2 concentrations going down. So, net, CO2 is not nearly the only cause of temperature changes. And, since we are now willing to do first cut approximations, we conclude in practical terms that, under anything like current conditions, CO2 is irrelevant. Bluntly, the variations in CO2 concentrations don't match the variations in temperature worth a darn. (2) The stuff I outlined about sun spots fit the temperature data much better. The example that sticks out like sore thumb is the Little Ice Age -- that is, the Maunder Minimum after guy Maunder who noticed that there were nearly no sun spots during the Little Ice Age. Otherwise, so far we don't know even as much as dip squat about the change in temperature, from CO2 or anything else, in the next 100 years. Clearly the first cut, safe bet is that there won't be any change. And, it helps a lot that currently the temperature and the changes in temperature don't look at all unusual. I.e., it was hotter in the Medieval times when people were growing grapes in England. Or, worry about if going to get cancer and f'get about CO2. Sure as heck f'get about carbon taxes.