I am enthusiastically in favor of increasing turn-over in democratically elected bodies, generally, but I think that the focus should be mobility rather than representation, so I don’t really like either of the original proposals.
Instead, I would suggest that elected positions should be encouraged (but not necessarily required) to mentor at least one project member who could reasonably be expected to advance into the elected role with the assistance of mentorship.
I would hope to see elected members have periodic one-on-one meetings with their mentee(s), in which they discussed some or all of: the work that the mentor has done recently as a function of the elected position, work that the mentee might undertake that would be useful to the project and demonstrate technical and social leadership, questions about the project or the mentor’s role that the mentee would like to learn more about, and feedback about challenges that the mentee has in the project (or challenges they hear about in the communities they participate in) and other opportunities for improvement, and (where applicable) how to present the mentee’s contributions and ideas in future applications for election.
I believe that term limits are probably not necessary in systems where elected members have confidence that there are other members who could competently fulfill their duties, and that elected systems should develop that confidence.
Fedora does a lot of technical stuff very well, but I think that a volunteer project is inherently social in nature, and it would be beneficial to focus on some of the social aspects of the project. Mentoring as a solution fits well with the “Friends” foundation of the project, and with existing documented mentoring practice.
Mentoring serves the specific purpose of getting newcomers into elected positions, and serves the broader goals of making the project more sustainable, and more attractive to newcomers.