Re: So, if they can't replicate anything...
Maybe. Being replicable is just the most solid form of science, and it isn't actually the value of the science. The value is in it's predictive capabilities. Having deduced Kepler's Laws of Motion we can predict where an object in orbit will be found at a different time. Or, seeing a deviation from where something is supposed to be, find another object which we hadn't noticed before.
You should be able to test that predictive part of the models. But that's where the problems start. All of the models predict far greater warming than we've actually seen. And you can't actually look at models, only the inputs and outputs of the models. The rest is "proprietary intellectual property." Hell, even some of the data is "proprietary" or at least how they have calculated the inputs from the raw data is. There's big money involved in all of this. Here in the US it's large chunks of some of the budgets at NOAA whether at the weather service or the division working on environmental predictions. Beyond that there are commercial companies making big bucks from processing and packaging that data. Some are just repackaging the weather data, others make money telling farmers how to improve their crop yields. So anybody claiming they have no monetary interest in this is full of crap.
Some of the monetary interests need solid data - crop yields would be a big one here. Insurance companies are probably another. Even the weather service proper needs good data for short term predictions. If you get 7 out of 10 hurricane or flood warnings wrong, people won't take them seriously, so those bits need accuracy. Others, like what the some mean temperature is going to be 50 years from now? Not so much. There's plenty of time to correct that. And here's the rub: with 50 years to correct it, your local Congresscritter is likely to cut the budget for it because it doesn't affect his election chances next cycle. So if you want that budget money THIS budget cycle, you have to create an emergency on which he MUST act NOW. AWG is just that kind of emergency. And the professional unelected politicians who put together the budgets know that. Given a choice between two scientist, one of who believes in AWG and one of whom thinks it is a crock, they pull from the AWG believer to get the money. And thus the research becomes self-selecting.