DOE labs shouldn't be involved in fundamental science? That's funny, since LLNL was founded by Teller, who strongly believed that a strong fundamental research component was essential to the vitality of the lab. Why? Good scientists want to work on real scientific problems, not figure out the reason why some piece of plastic is degrading inside a bomb. At the same time, they must be expected to work on practical problems of importance to the mission. As in any labor market, if you want top quality staff, you need to do what you need to do to attract them. That means having some fraction of their time devoted to good, original science. The alternative is to hire people with no ambition, content to work on whatever is handed to them, in exchange for a paycheck. Sorry, you don't want C students in charge of nuclear weapons.