PoIR's main point is provably false in this case.
I see the situation like this. The authors of legal briefs and court rulings have enough of an understanding to feel confident they can write meaningful arguments on the topic. But yet they do not understand computers and software well enough to reach technically correct conclusions. The unfortunate result is legal precedents that do not connect with reality.
A simple google search revealed that the attorney arguing this case before the Supreme Court has a BS in Electrical Engineering from Duke, and a MS in Computer Science from Johns Hopkins.
And he is arguing in favor of patent-eligibility for Bilski's method (which isn't even tied to software, remember).
True, the people *deciding* the case is a different story. The Supreme Court's decisions do indeed reflect a lack of understanding. In Benson they obviously didn't understand what a re-entrant shift register is.
But the problem here is not as simple as "lawyers don't understand computers." (no matter how much the geek community would like to pretend that it is). Its just not that simple. I've spent a lot of time thinking about this issue and there is just no easy answer.
FYI - I'm a patent attorney with a BS in computer engineering and my main practice is computer and software patent prosecution. I've programmed in Java, C, Assembly, etc. I've designed a simple pipelined (5-stage) CPU. I know what software is.
PoIR also claims that "All software is discovered and not invented" and that this is true "without a shred of doubt."
This unsupported claim begs the question - where is the line between discovery and invention. To posit that software is by definition NOT invented, and to conclude that software is not inventive, is obviously circular reasoning.
PoIR then goes on to ramble about Godel's numbers, and Turing Machines. While neat, these really have nothing to do with the issue.