Modeling studies consistently prove that either very few scientists are capable of creating accurate models, OR (and this is more highly likely) creating accurate models is entirely NOT the point, because accurate models would not yield the desired result.
Actually, those who understand models well also understand that the output of a model is 100% dependent on the input to the model. Freeman Dyson, British scientist, understood this when he pointed out that all climate models are rubbish, as computationally correct weather models require scale down to well under one mile, which computationally is impossible to do. Our computers are not, and cannot be made fast enough for turbulent, chaotic systems in four dimensions.
Another way to explain this is to note the smallest possible computer that has the computational power to predict weather (and then only for up to 10 days maximum) are - either Earth itself, or an exact replica or Earth itself, also revolving around this Sun.
The climate ‘models’ all use fudge factors. Reason: when they didn’t use fudge factors, the Earth’s temperature would increase by about one degree per day. The fudge factors allow the model to have an artificially induced stability that calculations do not permit.
In other words, the basic models are flat-out wrong, and the creators of the models know this. Further, you or I couldn’t make a model significantly better that used computation, as we can’t model climate due to the small-scale effects.