Is this a bit off-topic for us? No, since I’m aware that many of our readers are students, teachers, researchers, and the like. Simon Peyton Jones of Microsoft Research delivers a fantastic lecture on how to write a research paper that could be easily adapted as “How to Write a Great Book,” or maybe even a great blog post.
From the abstract for his discussion:

Seven simple suggestions: don’t wait — write, identify your key idea, tell a story, nail your contributions, put related work at the end, put your readers first, listen to your readers.

Some of these are things that as an editor I’ve been telling writers for years, especially the point about how, whether in a book or an article, ultimately YOU ONLY GET TO SAY ONE THING. Or as he calls it, a “ping.” Not two, three or four. If you have four ideas to convey, write four books, four blog posts, four research papers, etc.
Also, the idea comes first, followed by the writing, followed by the research.
However, he leaves out these two rules of thumb: Anything can be written at any length. And: With art and sympathy, anything can be explained to anyone.

I’m now on Twitter. Find me @d_klinghoffer.