I have to admit, I did not look for Feyerabend's books hard enough. But friends mentioned him a few times.

My first impression was that Popper was quite a positivist with the falsification method. On the other hand, if falsification is (or were) all that there is to the scientific method, than science indeed offers not so much, in a sense.

But Feyerabend goes too far as well with anarchistic or democratic views of science. Since most people don't even want to know basic scientific notions, it would be strange to let them decide what is science. Just as martial arts dans are not decided by popular vote, scientific standards deserve protection.

Of course, there are at least two sides of scientific research. For generation of new ideas indeed "anything goes" - if only you come up with a great idea, it does not matter whether you got it "methodologically". But then comes the work of testing ideas - and here we may discern a universal method in Popper's falsification. That method may not be complete - there might come new methodological framework, perhaps incorporating fully theories dealing with rare events, or differentiation between chaotic and systematic or cybernetic phenomena. But it is not "anything going" in the methodological evolution.

Universal distinctions of science and "non-science" may not be complete - Godel's incompleteness theorems are hiding somewhere behind (just as in Popper's falsificationism). But so what? Not only perfect knowledge, but perfect security or justice or government are not possible. Western philosophy is probably too much obsessed with universal rules and "fear" of exceptions to them. (Do we not see where insistence on perfect and most efficient markets and even democracies lead us to?) Scientific understanding and standards were evolving till now, and they will evolve in the future. Scientific revolutions will happen, but boy, you have to do a lot of work to make one. I doubt whether we need to be more efficient there.