I made that connection as well, but I wanted to present Dr. Seligman's work as directly as I could without editorializing it and then move on to discuss my own opinion of how it relates to hacking and programmer culture.

He may have had extremely good reasons for maintaining the subdivision between the three axes that are too subtle for me to grasp. For example, it could be that when writing tests to identify optimism, he found that he could only get a strong correlation between score and behaviour if he included questions from all three groups.

I don't know :-)