Perhaps all it takes is practise. This time, I spent twenty minutes reading my chapter and making notes, five minutes selecting which notes to polish, and another ten minutes reducing 125 words to precisely 100.
Today I submitted some corrections and suggestions to a security researcher in an attempt to help the security community by providing some editorial service. We had a fairly long discussion in the #pauldotcom IRC channel this morning about the editorial process. I am convinced that one of the answers to Richard Bejtlich's post about why academics ignore security professionals is the apparent lack of editorial control and process over the publications of professional security researchers. Hopefully I will be able to provide this service to others and help to prove my point. Somebody said that Maynor's article regarding his Apple wireless exploits was pretty much the current state of technical writing. (It was that article that spawned my rant.) Without going into the technical particulars of that paper, it's pretty obvious that the publication either lacked an editor, or the editor was asleep, as there's some pretty bad phrasing and at least a couple instances of horrible spell-check laziness there.
I don't claim that my writing is perfect - far from it - but I do claim that getting a second, third, or fourth opinion on a paper before it succumbs to a few thousand or more page hits is not a bad idea. If anybody wants some vicious commentary on the state of sentence construction or even just spelling in their paper, I'd be happy to provide it, keeping my own temporal restraints in mind. My own technical chops are poor, but I would like to think that I know a few things about writing well. I often claim that my honours degree was built less on solid research and more on clear writing.