Concerning the problems of planning and scale, why not just let it start very simply and develop its own course in a way. Some greate works of art (aka worth heaps) were made when the artist was blind -=- note: picaso -=- the guy had to be off his face -=- As a series on how to get a base system together It could act as a kind of wrapper for things like Gentoos' 1st stage and the LFS project. Becoming a form of template. There are also a lot of side effects involved which effect using Linux generally. In terms of compilation and cli/scripting techniques.
As a number of site people have put together their own LFS, and there are Gentoo people, including Slackware people ... there should be adequate forum feed back available, unless it's just ignored that is.
I'm suggesting to not so much to start out to create a distro as such ... that is quite ambitou, and does bring in the concerns of sustainability. Especially in its reflection on the mag if it just folded. But just a simple series on the workings of a base system. On the interalationships involved, why so much is apparently needed to get the thing to boot minimally in the first place. And that wouldn't be that simple really. Quite a deal of time could be spent there, especially if people did take it up to experiment with, which would be a good sign. There are sure to be lots of "how does that go again and why so" posts [ especially from me (grin) ]. Which would help in giving an indication on at what level, or in what way it should develop. If it does seem to catch hold then it could be gradually taken up to other levels. That would help to simlify the overall planning concerns, as it would then become incremental, a bit like an incremental system update ... It would also simplify the issues of commitment. There's no point in organising a party, just to find out everyone has changed there mind and gone out to a rage instead
Purpose ?, Linux started out as an enthusiasts interest ... so, why not. It would be extremely educational if not anything else.