I think the author has a point in saying that learning Git by trying to map it to Subversion is not the best way to do it, but I don't think analyzing it as a functional data structure adds much insight. To me, it is easier to understand when you look at its purpose, and how it solves the problems of that ___domain - and the biggest difficulties of version control are on account of the problem being essentially one of distributed, lockless concurrency, something not mentioned in this article.