> Git doesn't care about files, but changesets so this should raise a red flag with anyone with even a passing knowledge of how Git works. This is either a fundamental flaw in the understanding of how Git works by the OP or intentionally misleading wording.
Git absolutely cares about files. It cares about commits too, but a commit is mostly just a set of files with a pointer to its parent. The commands being discussed operate on individual files except for commit, which operates on the entire index (which is, conceptually, a snapshot of file).
> In the Commit section, the OP writes: It then points the current branch to this new commit. I don't like this wording because it makes it seem like a branch is something other than a reference to a particular commit (which is available in .git/refs/heads). Really, any reference to "current branch" should be replaced by HEAD, because the current branch is nothing other than HEAD, and HEAD of course is just a commit.
I don't understand your objection to this phrasing. A branch is a pointer to a commit; the statement talks about changing which commit the branch points to. I don't see where it implies anything else about what a branch is. As for "current branch"/"HEAD", using HEAD is more precise in terms of the actual implementation, but "current branch" is just as accurate and more (new-)user friendly. The author does discuss the relationship between the two prior to that point: "The current branch is identified by the special reference HEAD, which is "attached" to that branch."
> In the Checkout section, the OP writes: The checkout command is used to copy files from the history (or stage) to the working directory, and to optionally switch branches. This is blatantly false. `git checkout` doesn't "copy files"; it may move HEAD or it may get rid of any changes in the working directory, and in doing so may change the state of the files within the local repo.
When used as "git checkout <ref>", this is correct, but that is not what the author is talking about. "git checkout <ref> <file>" or "git checkout -- <file>" does exactly what the author describes, copying files from the given commit or index to the working directory.
> In the Committing with a Detached HEAD section, the OP writes: Once you check out something else, say master, the commit is (presumably) no longer referenced by anything else, and gets lost. This is only partly true. The commit still remains in the reflog and can be retrieved up to the point that garbage collection is run. It's incorrect to tell someone their commits are lost as soon as they `git checkout` another treeish.
Only partly true, but true enough. The commit isn't irretrievably lost, but the reflog falls more under "advanced usage"; certainly a git beginner shouldn't be making detached commits and relying on the reflog to get back to them.