Think about it at the molecular level.  If the AA metabolite is very reactive, that's very bad.  LTB4 has been described as an incredibly powerful chemoattractant, for example.  My guess is that the "good" metabolites are either just not as bad as ones like LTB4, or else they are like Mead acid metabolites, and they are less powerful because there is better antioxidant protection (for whatever reason) and/or less oxidative stress.  All this is contextual; in other words, if you had a gaping wound and bleeding profusely, you might want LTB4.  However, I've found that my nosebleeds now clot much quicker than before, or else it may be that the clot is rubbery, and so does not break up easily.  In order to know for sure, all kinds of experiments would need to be done, controlling for all these molecules, and that's not likely to be done any time soon.  In the meantime, one must look at the evidence and try to make sense of it as a whole.