I was a chemist by education and worked as a chemist from 91 to 99 before leaving for IT. I took a job as environmental database manager at an engineering firm. My family had relocated to the DC metro area for my wife's work and I had trouble finding a good job as an experienced BS level chemist. I have actually been a programmer \ developer now for most of my it career. Never looked back, never regretted move, and certainly making way more money than reasonably could expect to make as chemist.

I do not want to work for the insurance industry. If my company asked me to I would try to get out of it but, I would not leave my job over it. There are plenty of other areas that I would not get involved in as well. The question is hard to answer, it's not simple black and white. A blanket yes or no type answer does not really cut it. I would not write the spy program. I would work for the industries that you listed if the employor where legtimate.

There is a lot of discussion here that seems to be more about individual learning styles rather than the best SE books. Personally, I get a great deal from reading. I always have. However, I would not list anything as skill on a resume from 'book learning' alone. I would not feel comfortable claiming to know an SE methodology, language, pattern, or whatever unless I had actual developed something with it. My own code. Not just running samples. I feel that SE books have had a greater impact on the overall quality of my code than have the nuggets of code and advice I have gleaned from the internet. OOD/OOP concepts, Design Patterns, database design big picture things all come better to me from books. Remembering particular languages switch statement syntax or looking up sample implementations of a pattern is where the internet comes through for me. On a daily basis for the little things I hit the F1 key, then if necessary google, then maybe a reference book. But when experimenting with a new larger topic like learning a new language or development methodology I hit the books first. After reading a bit, I may actual drop this subject from my interests or pursue it in some actual coding effort. If I don't actually use the new knowledge on something I will likely forget most of it fairly quickly.