1. The guide is intended for people who want to take a look "under the hood". This is really difficult to do through GUIs, which tend to hide all of the implementation details.
2. As a child comment mentioned, even if I include GUIs in the guide, I can't possibly cover all of them. This just confuses things for people who use GUI X when the guide only talks about GUI Y and Z.
3. It's much easier to write the guide using CLI examples. I wrote this over the weekend; I don't think I would have completed anywhere near as much of it if I had to keep taking screenshots, cropping them, uploading them, etc.
Besides, there are already plenty of other guides that talk about "how to use Git", "how to use GUI X to do Git feature Y". I wouldn't want to detract from the main goal of this guide which is to give people who are already comfortable with Git that extra push to "get" how Git models and interacts with your repository.