I try not to write about things I don't understand, so I've avoided taking much of a position on global warming. Much like Sam Cooke, I don't know much about Geography or Biology.
And while I don't know much about the Science Book, I do understand some fundamental principles of science. One of which is that experiments should be reproducable.
The Climate Research Unit has not only refused to release data under the Freedom of Information Act (due to confidentiality with other countries, but they won't say which) but now say that they don't have the original 1980s data from weather stations, just "value-added" information.
So in other words, there's actually no way to verify any of their experiments, except to trust them, which frankly means that whatever they've produced as results belongs with perpetual motion machines and religion.
As this is all done with computers and models, it should be perfectly simple to provide the raw data that went into the model and the source code of the model to allow for full interrogation. When they do that, I'll start taking this seriously as science.