Hey man, Mondongorongo here; i noticed you linked my blog up there, thanks for that.
Though I blog in spanish and that can be kind of difficult for some people I'll work on a translation of my future blogs in Gamasutra too(maybe in french too, haven't decided yet). I've been doing this for a little more than a month, and it has been great as a learning experience, you have to work hard and stay focused on your deadlines, I've devised some methods that help me accomplish my goals in one week. Let me copy and paste it from here
I start the first day by grasping the idea of the project and doing a conceptualization through rough prototyping that gives me a clearer idea of how the project could unfold, after that i write down my main concept, core mechanics and key features and the expected deadline of the project; i then spend the next 2 or 3 days in sprints of development, that sometimes give me additional features that i prototype and test but if they don't add to the experience or don't seem achievable in the timeline, i toss them out. Finally i go through a couple pf days of playtesting and polish where i balance the opinion of others with my own and see if some improvement is viable.
This gives me a beta that i release to the web(although it's mostly game prototes i call them "betas of prototypes") and one or two days later i postmortem the project.
This has worked in these small scale projects, but i think that the core time management and feature validation process is a good foundation for when i need to scale onto bigger projects.
Besides my delusions of grandeur(i just quoted myself, for crying out loud XD), it's awesome to get some sort of feedback on your games; playtesting is very important, as is the shameless self promoting of your work, but also you have to develop a thick skin and learn not to take it personal, sometimes great ideas and good intentions end in bad product and people can be harsh, specially if your tersters are friends and family as is in my case.
So yeah, just wanted to pop up and give my thanks and my two cents.
*prototes in mondongoronguesque means prototypes 
EDIT:
I've never heard of a game a week or every two weeks. I've only heard of jams and the one game a month ( http://onegameamonth.com/ ). I'd think that the failures you get with one game a week or two would be the same as the month long one, but would likely run the risk of you getting burned out faster than most programmers.
That's not 100% true, they are somewhat different; you have to understand your limitations and know that you can't aspire to build the WoW killer in thosae 7 days, but it helps a lot to develop a kind of lean approach to design that's very interesting; not only that but it also helps you to focus on identifying your themes, key features and concepts clearly and faster.
Also, and paraphrasing, Rami Ismail, the process of releasing a game is sort of a muscle itself, you have to flex it to develop it, and i flex it 4 times in a month instead of 1; don't get me wrong, eventually I'll scale to a game a month, or a game every three months, then onto bigger projects, but as I mentioned before, you really need to develop the thick skin, the ability to promote your work, to approach testers, to ask for feedback and criticism. It's crucial for the learning experience, as developers we tend to focus exclusively on the tech side of things, but a greater understanding of the whole process has a great educational value.
Game a week is not done to create the greatest game ever(though if it happens that's great) it's more about flexing the game development and game design muscles needed to become a better developer. It can be burning for some if you strech it too long(kinda happened to me this week), but if you can use it correctly it's a great tool for starting game devs.
