From your posts so far it looks suspiciously like the summer of code was not managed as well as it could have been and, with one or two noteable exceptions, I see little evidence students delivered long term contributions that significantly benefit the project.  Although this was google’s money and not Debian’s, I wonder if good value was actually obtained here.

How will Debian learn from this? –  I do hope that this summary will be taken further than just your blog.   (Although self-proclaimed “geeks” are often suspicious of overly beaurocratic large “coporate” organisations, it’s always struck me that such organisations are much better at learning lessons than Debian, where good ideas are raised, improved, half-heartedly implemented then abandoned.)

Here ar some other questions that come to my mind, although I suspect you already plan to consider most of them: Did anyone have a “high-level” role in monitoring the projects?  Would such a role have helped?  Were problems noticed on time?  What was done to resolve them? Why did so few students remain active in Debian after the summer ended?  Why are so few projects actually used today?  Were mentors and students who did deliver appropriated praised? What about those that did not deliver? how did google assess Debian’s management of the summer of code?  How will Debian change their approach for next year?