As the author an maintainer of several small FOSS projects, I rarely receive pull requests that I can merge right away. In that sense, a distributed VCS doesn't really improve over the patch-by-mail approach for random contributions (it does helps cooperation among regulars however).
I can also tell you right away that a FOSS project without direction or maintainer quickly dies, independently of the VCS. Successfull forks are very rare. Mostly, projects with good potential but no maintainer get simply dumped and reimplemented by another programmer.
If you are a maintainer, you will appreciate all the points the authors is making. If you are making a contribution, adhering to all the points will increase the odds of your change to make it quickly and improve the software.
Also, conversely as a contributor, I don't spend time on projects which have no maintainer or lack direction anymore. Having the source of a project (especially large) sounds useful at first, until you realize that maintaining it without the inner knowledge is a daunting task. It only makes sense if you are willing to use it extensively (and become the new the-facto maintainer).
Of course, the articles misses an important point: talking to the maintainer/author first. If you were about to do a big change, would you know if there were any chance to be integrated first? Maybe knowing if someone is already working on it? (heavy refactoring comes to mind). I've gained precious insights about design by simply talking to authors before starting to work on a contribution, saving my time countless times.
If your intention is doing a change no matter what, then you're not really interested in _contributing_.
"The maintainer is in for the long-haul, and has been working on this code for longer than you have. Chances are, he will keep doing this even after you have long moved into your next project."
That might be a common pattern but it seems like it's sort-of a dead end pattern. If put yourself out there as The Maintainer in that way, you've basically said no is going to be able to make major contributions without you - going it alone become a self-perpetuating process and it's indeed not surprising that no one will stick around.
I have a large project that I hope to release in the next 1-3 months. I've put a large amount of effort into it. I'm also putting effort into making the code good enough that others can come on board and make significant extensions to it. Following the existing code style certainly but quite possible rearranging functionality, breaking up and rearranging function if it's useful and so-forth.
A VCS like git that lets everyone have the branch they like can be good here. Do what you love and inclusion in the main branch should follow sooner or later...