An interesting read, my only issue, as with many of these papers is that it is preaching to the choir. Well at least within the realms of the POTN RAW and post processing enviroment. It would I think make a much more informative read for many if posted in the General Photography forum. The biggest issue there is getting enough people to actually read all 17 pages.
I do like the way he tries to show that a JPEG file sooc is a finished product and so the equivalent of a print. Although as one who learnt photography with film I feel that the JPEG is more like a transparancy. If you are going to project it on a screen, then it's indeed the finished product. But I could take a transparancy into the darkroom and further process it to make a Ciba-Chrome print. In much the same way you can do far more to a JPEG file than you could actually do with a physical print.
I beleive that the RAW file is actually much more like the latent image held on an exposed but unprocessed peice of film. The camera's sensor will have an effect that is much like picking your film choice. The RAW processor is like picking your chemistry. The great advantage with digital is that the first stage of processing, unlike film is not necessarly fixed, so we can move back to it if needed to. Also the improvements that we have seen in RAW processing software since the paper was written (it seems to have been published when CS3 was current) now also means that we can selectivly process different areas of each image, thanks to local brushes, something that is just impossible with film.
Alan