What is the opinion of pure economists on the economics modules of integrated assessment models?

My background is dynamic systems modelling (primarily aerospace and power systems) and I'd like to get into long-term macroeconomic simulations. It's not strictly necessary for my current job, but I feel like it would add something to my portfolio. I know that not having an economics background and trying to model and simulate economics is sure to provoke some people, but please give me a fair shot.

Anyway, when I say "long-term" I'm talking about the next fifty years or so.

I have read publications on integrated assessment models, but – without being an expert on economics – I get the impression that their economics modules are sort of worthless. Old, pre-IAM work such as world3 suffers from even more crippling shortcomings in its model behavior.

I suppose to start with my question would be whether economists consider IAMs state-of-the-art in terms of economic modelling (leaving out all the other modules for now). As an addendum, if they do not greatly trust IAMs (and this is my impression based on most of the papers I read), are there better models available and why are they not used? Or do economists broadly feel that such time horizons can't be simulated to any credible degree? If the dynamics of the world economic system are too sensitive to model and parameter uncertainty to allow a trustworthy forecast of a couple of decades, this should be relatively simple to demonstrate and make further efforts to refine long-term IAM modelling futile, but the efforts continue unabated.

I feel like I need some guidance and context on these things. Thanks!