Examples do not prove anything
And this is what this seems to be. Just because sales for EMI didn't rise when they ditched DRM doesn't mean DRM isn't bad for business. Example isn't correlation, and correlation isn't causality.
There are many counter-examples, at least equally valid:
For example look at the early modern German book market as compared to the British one. In Germany there was no strong Copyright, so books could easily be copied. This lead to many people learning to read, and books becoming cheap. This also caused a lot more books to be published and more money for the writers. As opposed to Britain where only the publishers got rich.
Also it's missing one point completely. The cat is out of the box. We do have uncensored high speed data networks. People are going to "illegally" get what they want. So what now should be done is finding ways to deal with it. Just like the industry found ways to deal with automatic pianos, the radio or even Youtube.
People usually don't pirate in order to save money. They pirate because they cannot get a paid copy. Many movies will probably never come out on DVD, for example ET (1982, not the 2002 version which is horrible and currently the only available version). Some people cannot afford to pay $20 for a DVD. Also I'm sure many pirates would lovingly give twice of what the artists get from buying the content directly to the artists, but there simply is a fairly useless industry in between.