IMO, the hard science underlying climate policy has been extremely narrowly based, and thus not really credible or worth taking terribly seriously. The glaring problem - pardon the pun - is the omission of solar forcing (other than radiation) from the models. Wonderfully, this situation is in the process of changing! As of the 2022 IPCC report (for the first time ever) solar particles including solar wind, flares, and CMEs, as well as galactic cosmic rays, are to be taken into the natural forcing model. Then, and only then, will we have a comprehensive and hopefully credible insight into climate. As these additional insights come into consideration, I feel climate science is going to undergo revelatory changes and improvements, and will ultimately begin to enable meaningful and constructive policy from regulatory agencies.
I understand your position and I can sympathize with it. It is exactly the kind of think I was complaining about 11 years ago in the post I linked.
I used to do very similar analysis to what the climate scientists are doing. In fact I attended lectures by colleagues who were working on that that exact subject because our disciplines were so closely related. My particular problem was the problem of mapping out spacecraft trajectories. I modeled complex unknown systems for the purpose of determining where a spacecraft was. I did this for two of the most complex systems in our solar system (I'll stop short of naming the space missions, but they're well known). I'll give you an example that I'm just going to draw out of a hat here.
The Saturn system is a very difficult system for determining where spacecraft are. You don't know where any of Saturn's moons are, and you don't know the rings very well. Saturn itself is pretty well known, but a lot of that is coupled with the moons, so teasing out the difference between the moons and the planet requires some time and data with which to differentiate the signals between the two. Gravity fields are far more complex than just "how massive is it" too. Various numbers of terms are used for various gravity fields, that goes for Saturn and saturn's moons as well. You don't have a great idea where the moons are, let alone how massive they are.
So if you have a spacecraft that is bouncing around the Saturnian system, it gets exposed to all of these effects, solar radiation, distributed ring mass, distributed mass within saturn, distributed masses within saturns' moons, whose gravity you do not know. And then perhaps that spacecraft goes near Titan and experiences atmospheric drag. You don't know the drag coefficient of your spacecraft because it wasn't tested in a wind tunnel. And of course the spacecraft itself is not particularly well known. You have outgassing and radiation forces from the craft, but also the thrusters (attitude control and main propulsion). And the thrusters degrade over time, and you don't know exactly how they'll do that.
All of these variables for this very complicated system go into a giant model, with uncertainties for all of them (sometimes very large uncertainties). And then you get just a few measurements. Range, and Doppler along a straight line with Earth, and also optical navigation images which are taken of the various bodies against a star field. If you can figure out where the spacecraft is, you can use images of the moons to help determine where they are. Not a lot of measurements, and those measurements can be incestuous, constantly measuring the same thing but never giving you insight to bring some other kind of uncertainty down.
You accumulate data and beat the uncertainties down, slowly, and refine your model. And that's how you do a flyby to within a few dozen or so kilometers of the surface of a moon without trashing your multi-billion dollar space mission.
This is exactly the kind of model that the climate scientists use. They have many variables, modeled to various levels of fidelity, and with differing uncertainty, sometimes drastically different. It is the same math, the same underlying problem, but one which can benefit from lots of different kinds of data. They add data from many different kinds of sources, and go back as far in time as they can trust (with a healthy dose of uncertainty of course) and try to beat down their unknowns. The models they have been using have been increasing in fidelity rapidly over the last decade (and the decade before).
Your complaint about modeling solar flux, cosmic rays, etc. is a valid one. And it's a hard problem, but rest assured it is one they are acutely aware of and that they work on. The thing is, the entire model is a tuned system. Tuned by decades of data. You can add a variable in and see all of the ways that it soaks up a signal from something else, and see whether or not it "fits". I did this all the time. We even had models for variables that were entirely for soaking up missing signals, and when those started stepping out of line, you knew something was wrong with your model. You could then look for higher fidelity terms that allowed the soak-up terms to settle back down. This process is how the model grows in complexity. Something doesn't fit right, something moves by 3 sigma, soak-up terms start to move in a big way and indicate that something is wrong, something is not modeled properly.
The introduction of the types of functions you're referring to will allow the model to adjust certain kinds of signals, but not others - because there is underlying physics for how they must necessarily interact with Earth. The more mature your model is, the more you understand whether these terms are even
capable of moving or driving the net result. It can become very apparent, very quickly, that you simply cannot model the system properly without taking into account CO2.
What climate scientists have been doing over the last decade is moving toward a convergent model that has been shown to be accurate against the data we had, and also accurately predict future data. This is the acid test for modeling of systems like this - how well does it predict the future. That's what I was looking for 11 years ago, and it has been demonstrated. This scientific field is new, but it has been growing rapidly and it grows on the well-understood mathematical underpinnings of estimation theory. I'm not saying they have this thing locked down to within tiny uncertainties, but I have become convinced that they at least know what they're doing well enough to reach some conclusions. The conclusion is that they simply cannot fit the data without human-caused greenhouse gas warming.
