Well -
Talldave - your argument (well - Armstrong's argument) is easily demolished.
Armstrong (who wrote the source paper from which you and Wattupwiththat are quoting) is an economist, with track record in econimic forecasting. This is stochastic - and has no physical laws that determine result. Human hopes & fears are difficult to forecast - and it is well known that experts tend to get them wrong.
Replace "scientific" by "economic" and I would agree.
Armstrong - in applying his economic principles to climate change modelling - therefore not surprisingly misunderstands climate models and thinks that they are wholly stochastic. Ecomomic models are a matter of getting best fit to data, physical laws don't exist. Also he clearly can't distinguish between climate and weather and sees future climate as being a chaotic process like future weather (whereas when spatially & temporally averaged of course deterministic trends emerge from chaotic noise). And he has neither the expertise nor the inclination to read the literature that he dismisses.
It is as though a fisherman with 30 years experince catching fish wrote a paper attempting to dismiss evolutionary theories about how mammals developed from fish without bothering to read the extant literature.
Climate models do have parameters which must be determined from fit to data - but these are at many different levels and the data that fit them are not the global temperature that is the forecast/hindcast. That global match provides some extra validation (in case of hindcast). Other data: mesoscale, micogrid modelling, is used to determine the vertical mixing parameters which are impossible to model directly over long time periods and the whole globe.
This whole structure of physical equations, modelling, validation is foreign to economic forecasting. indeed only somone who had no understanding of physics would confuse the two.
For a vicious (not really to my taste - but "tit-for-tat" application of "scientific principles" to Armstrong's own work on IPCC see realclimate). I can see why the pro-AGW web advocates like Gavin reply to character slur and partial reporting with their own material of the same kind, and to be fair, they do not do this stuff often.
http://www.realclimate.org/index.php/ar ... -forecast/
Simon has a more subtle line of argument. It has two prongs:
(a) GCMs do not (cannot) model NS equations well enough to have any predictive power.
(b) Past 30 years apparent global warming is explained by some combination of (select your anti-AGW argument of choice - many of been advocated - none have withstod scrutiny. Nowadays few get into the serious scientific literature and when they do they are rebutted).
(a) Is subtle and difficult to challenge. Neither Simon (unless he has written papers of a standard that would pass peer-review) nor I have looked at the field enough to be authoritative. I take the view that 1000s of peer-reviewed papers of 10s of years will expose obvious holes. especially because there are many academics who are emotionally wedded to the "AGW is rubbish" viewpoint. Anyone can publish papers and they will be taken up by others if ideas are good. You might imagine that climate modelling is a closed coterie of like-minded advocates. Read the IPCC AR4 report - there are a large number of models with warming predicted between catastrphic and very mild. As time has progressed the models have reached greater agreement, though there is still a wide variation. And the average has been moving towards greater warming.
So the evidence we have on this (most subtle and difficult to debunk) argument is really meta-evidence.
Do you trust the scientific peer-review process? (Not do you trust the scientific community!).
Do you find it strange that if the holes in AGW are as large as your favourite AGW blog claims, there are NO serious papers knocking down AGW which have been published and not comprehensively rebutted? There is diversity of opinion among both academics and editors of journals.
Note BTW that the answer to a) shifts with time. A few years back GCMs could be criticised for not modelling ocean currents - now they do. The computing power available P, and therefore just on the basis of brute-force computation the spatial resolution D (D ~ 1/P^(1/3)) gets better.
(b) is frankly feeble. The anti-AGW arguments (bad data - heat effect - variable solar forcing - ocean current dependence - etc, etc) all fall down when examined. They all HAVE been examined. And anti-AGW advocates have shifted ground from "GW is not happenning" to GW is not anthropogenic" or "GW is a multi-decadal effect".
Unfortunately - as proved in this forum - these anti-AGW myths, rather like the fertility ritual mythologies of ancient societies, have great staying power in the uncritical blogosphere.
Best wishes, Tom