As for me, this model seems to be the natural choice for branching in presense of multiple parallel development lines and many releases. Unfortunately the article has a couple of weak points. First, it is written in a horrible style of “first we discuss the obvious draw-backs of inadequate solution, never saying that it is actually not recommended; then we finally `discover’ the correct and simple one”. Why waste time?
Also, this article is written in somewhat dry and formal manner, but I guess this is pre-requisite for IEEE publication.
Nevertheless, this is worth a read, just skip the “Branch-by-release model” and “Build-by-bug-number syndrome” chapters and read them after you finish the rest.