I disagree. I believe it's the load-bearing analogy that's carrying all the ideas in the article. Consider e.g. the sections "The benefits of small docs" and "A

I disagree. I believe it's the load-bearing analogy that's carrying all the ideas in the article. Consider e.g. the sections "The benefits of small docs" and "Avoid blocking yourself during reviews" - they're mostly about workflow. They apply to changesets, not the artifact itself. I feel like there's an implied assumption in the article, that documents are always submitted and reviewed wholesale, so to get the benefits of small commits, you need to make the document small enough to qualify as a small commit. If that's the idea, then I disagree with it.