Then rigorously show how this is the case. It's really that simple. The fact it has not been done yet should tell you something.
I'm not sure of the 'this' that needs to be proven, but I'm going to interpret it as the GP statement scientists don't have the climate models rightas something that isn't nuts or crazy.
From Chapter 9 of the IPCC AR5, complete with more than a half dozen citations:
For instance, maintaining the global mean top of the atmosphere (TOA) energy balance in a simulation of pre-industrial climate is essential to prevent the climate system from drifting to an unrealistic state. The models used in this report almost universally contain adjustments to parameters in their treatment of clouds to fulfil this important constraint of the climate system (Watanabe et al., 2010; Donner et al., 2011; Gent et al., 2011; Golaz et al., 2011; Martin et al., 2011; Hazeleger et al., 2012; Mauritsen et al., 2012; Hourdin et al., 2013).
And later the IPCC goes on to note:
Model tuning directly influences the evaluation of climate models, as the quantities that are tuned cannot be used in model evaluation. Quantities closely related to those tuned will provide only weak tests of model performance.
So to recap, the singular driving force of all climate change, the energy imbalance at Top Of Atmosphere, is NOT an emergent property of the underlying climate models, but instead still requires hand tuning to avoid running out to an unrealistic state. More over, by the IPCC's own standards, quantities closely related to those tuned are only weak tests of model performance. How many components of the climate aren't at least weakly related to TOA energy?
It keeps going though, as if you hand adjust clouds or other parameters to balance energy, if your results were off you'd expect to get the macro of energy balance that was tuned correct on the mean, but because of compensating errors too high here and too low there. Read the entire IPCC chapter linked above and count the number of times compensating errors are observed in the unknown parameters like clouds.
If you further your reading beyond the IPCC chapter and read the linked journals you'll even find that climate models still regularly, as in more often than not, don't pass the conservation of energy test, it's even stated as part of the reason that tuning TOA energy is still necessary until bugs in code or algorithms can catch the leaking energy or additional energy that coming out of the ether.
All that said, climate models truly are still good tools. The steps taken above are still good steps, and I honestly and truly mean that. They all still contribute to testing theories and ideas of how components of our climate function and are vital tool to furthering our understanding. At the exact same time though, to declare that lacking confidence in future prediction from them today is nuts or crazy is just wrong. There's very good reason to place very big caveats and conditions on the projections currently being generated. In particular hindcast skill should NOT be expected to be a very good indicator of predictive skill at all.