Hi Mark,
We did, not enough and just a bit too late. ;)
A few people started the trend (including me), but we didn't have enough organization to propose a good number of interesting projects in time. Most of the comments was about the quality of the work and the time spent on mentoring being too much.
This year, I followed the work of some students on the LLVMLinux group and the mentoring time spent was very little. The results of other LLVM projects (ex. Flang) speak for themselves. I wanted to reiterate that the benefits of the GSOC almost always outweigh the costs, even if the project is not successful, but most are.
It's good to have corporate backing (ie. you ;), and I think we should start planning for the next year much sooner (maybe even during the US Connect) and have some incentive plan to foster mentors inside Linaro. It's a great way to involve the community in working with Linaro towards the common goal, and to recruit candidates in subsequent years with proven track with Open Source *and* Linaro.
cheers,
--renato