I replied to this post and your other one via the google group website but my replies did not show up in the conversation thread. So let me know if it didn't get through to you. (I'm replying now through email.)
Quick answer: that's an older help page from old website for an old release.
Current help page for that parameter is here and you do not have to externally generate decoy sequences even if you're using the old release as you're likely not searching an indexed database.
Quick answer to second question: use monoisotopic masses and 0/1/2/3 isotope_error (assuming you're using ppm precursor tolerances)