People usually admit guilt, and I don't think I have anything to do with those models, as some of those are older than me. Of course they were flawed. Models are always imperfect, but the important things is, are they useful?
Well. Ultimately, we shall see the usefulness of these climate models, and those old models were useful to show some the extremes, huge swings, big effects. Solar cycles, big non-sea ice melts (Greenland, Antarctica), no reduction in emissions, a lot of CFC (used for refrigerators, you probably remember them).
Furthermore, it's kind of hard to experimentally test the crude models, as you'd need to copy the universe, and do this or that in a lot of copies. Those models were simple, but still full of statistics, so the chances are still chances, just there are hundreds of millions of people who don't want to put Earth on a roulette table.
Hm, I never considered my worldviews as static. What suggests that?
So, now those are concrete scientifically explorable questions.
First, sure, the CO2 content of the atmosphere is increasing a lot slower than we're adding it, so it goes somewhere. Vegetation, oceans, and some could escape to space, but it's basically zero. So from a CO2 standpoint, Earth is a closed system ( https://en.wikipedia.org/wiki/Atmospheric_escape#Dominant_at... ).
There are thousands of people working on determining the composition of the atmosphere in the past. We have very good data for about the last half million of so year. ( https://en.wikipedia.org/wiki/Carbon_dioxide_in_Earth's_atmo... ) And getting better and better estimates for the longer timescales (hundreds of millions of years).
The main point of this whole global climate change/warming/alarm/denial is that humans are very busy and we are doing something very fast. Big quick changes. We released so much CO2 and CH4 (methane) and other gases, that it is having an effect, much larger than whatever would happen to any natural cause. (Volcanoes release yearly less than a gigaton of CO2 (about 0.3 Gt per year) and we have good data for the last 50 years, and there was no significant increase in CO2 after big eruptions. So, they released a lot during the millions of years, but that largely went down into the sediments at the bottom of the oceans.)
The Sun has many cycles, the short 11-year sunspot cycle is not that interesting, it happened in the last century ~9 times, and nothing interesting happened (<0.1% total irradiance change from cycle min to cycle max). The big cycles are the Milankovitch cycles (various periodic changes in the orbital position of Earth relative to the Sun), those are important for big Ice Ages, as total irradiance can change ~25%, but these are slow, take 20 000 year or longer. And the models are of course not perfect for what these orbital cycle things mean for the climate on Earth: https://en.wikipedia.org/wiki/Milankovitch_cycles#Problems
So, all in all, we released more CO2 in the last 50 years than "nature played with" in the last 400k yearly, as in we released a half a teraton (500 Gt) in total and the yearly variation was 200 Gt. Each year we release ~35 Gt, that has to go somewhere or accumulate in the atmosphere. (And as you noted, not all of it is accreting, hence all this text and discussion.)
(Furthermore, the average temperature change we're talking about takes into account the seasonality from the Sun, so the trendline is statistically separated from the baseline.)
Well, I treat the IPCC as the aggregation of climate scientists. (Best data, best models, most rational interpretations, and so on.) They are internationally funded (there's a breakdown of contribution by country, it's chartered by the UN, but the world meteorological organisation pays its yearly budget), so far they are pretty cheap, as of 2014 they received about 138 M CHF (136 M USD at current exchange rate). But probably a review of their work could be crowdfunded pretty easily. (Though we'd have to discuss what is a good method of estimation and what isn't, what is a good way to model this or that and what isn't, but of course these points of concern should be listed, and .. there is a list https://en.wikipedia.org/wiki/Criticism_of_the_IPCC_Fourth_A... ) After the IPCC's 4th report in 2009, there was a lot of scrutiny of their work, conflicts of interest and so on, and their process benefited from this, as the 5th report now uses better data, better sources, a more unified probability framework (so probability estimates are consistent with each other), and so on.
I believe that when thousands of scientists endorse a result, and spend time talking about it, spend energy explaining it, how it came to be, and why specific "counter" papers are flawed, then that result is okay.
The times are changing, I don't know what'll happen in the future after this mini-collapse of intelligent discourse, now that truth and English are basically arbitrary, now that Enlightenment values of clarity and rational attention to detail no longer matter, with the rise of the complex charlatans that are willingly or unwillingly put certain issues before scientific integrity. Now that more and more people are forced to make less and less optimal compromises with supporters of ideologies foreign to theirs.
But that's a bit too doomsaying and offtopic. Back to models.
The IPCC reports are based on technical models integrated into a bigger reporting framework. And those models were covered in a lot of papers, and the best models were chosen, and then computationally run. (See the Climate Model Intercomparison Project, also, "When using blended temperatures the observations lie well within the 90% range of RCP8.5 runs for the whole of the last decade" -- RCP8.5 is what IPCC used, and the blended things is about estimating air temperature (the models predict air temp) from sea temp (what satellites measure when they look at the sea) -- http://eprints.whiterose.ac.uk/95782/1/blend_paper.pdf )
That said, SCIENCE IS FUCKING SUCKS when it comes to presenting results. (Compare to "tech": https://areweideyet.com/ https://arewefastyet.com/ http://www.arewewebyet.org/ vs http://cmip-pcmdi.llnl.gov/cmip5/publications.html and https://cmip-publications.llnl.gov/search?type=project&optio... there's no one-liner. Yet the question is pretty easy "how accureate are models on some metric, and what does that mean".)
"I know a lot of false claims were made, [...]" .. I haven't said that. I know the models were simple, so crude, bad, inefficient, inaccurate are not surprising to me. Not to mention there was not enough computational capacity in the 90s.
My message is, logic is powerful, but that doesn't mean simple. It means we can say that there are more and more possibilities. Such that if someone is screaming more and more, it doesn't mean that they are just trying to scaremonger more, it might mean things are getting worse and worse. (That's how this subthread started.)