Hi,
I hit across this idea in the recent GSoC Mentors summit, and in the
discussion with Srishti and Sumit on the reducing usability and scope of GSoC/Outreachy projects[1] among the years.
*The problem* Students show up one or two weeks before GSoC or Outreachy, and propose a solution to existing ideas, and often end up completing it and leaving the project. Due to this, there is a decline in student-proposed ideas as well, given 1-2 weeks is not enough to understand Wikimedia from any direction.
I didn't really understand this. You seem to be talking about some or all of the following issues:
1) Fewer students doing projects for the Wikimedia Foundation as part of the Google Summer of Code and, I guess, Outreachy, than in previous years - 2013 being the high point.
2) Students doing projects that are less useful than in previous years.
3) Students not staying with the Wikimedia/MediaWiki after the conclusion of their project.
4) Students doing projects proposed by existing MediaWiki developers, rather than projects they proposed themselves.
I see these as four unrelated issues, and actually I see only two of them as real issues: #2 I don't think is true (though I'm not sure if that's what you meant by "usability"), while #4 I don't see as an issue at all. Personally, I think only projects proposed by potential mentors should be considered at all, and that the documentation should state that clearly. I'm not aware of any GSoC projects where the student came up with the idea on their own and then executed on it successfully - with the exception of projects where the student is an established MediaWiki developer who happens to currently be in college, but that's obviously a special case. It's just not reasonable to expect that someone from outside the WMF/MediaWiki community would be able to come up with a project that (a) makes sense, (b) fits within the current development roadmap, and (c) is of the right scope for a GSoC/Outreachy project.
More generally, I don't think there's anything less rewarding about doing a project that someone else came up with. In software development, as in most things, the difficult part - and the most rewarding part - is the execution, not the original idea. (There are various inspirational quotes to this effect.)
That leaves #1 and #3 - fewer students participating, fewer students staying on afterwards. I think #1 is just a function of fewer potential mentors getting involved. Retaining students, on the other hand, is a real problem. I can think of various ways to try to improve this, though creating a new outreach/funding program seems extreme - it would take a lot of work, and you would presumably run into the same problem of a limited number of mentors. If there's money to pay for these kinds of things, why not just put more money into, say, hiring more developers from out of the GSoC pool? It's one idea.
-Yaron