Carol, I think you're missing something important here. Aside from the fact that this would cost about $2 million a year, the structure you are proposing would only be providing support for English Wikipedia. (That is a lot more than the budget for the entire global Community Advocacy department. Just to keep things in perspective.)
Now there are a couple of questions to consider here. The first is: What problem, exactly, are we trying to solve? Next would be: is this a problem that is endemic on all projects? (I think we all know the answer is "no".) So then we move to: Where all is it a problem? Large projects? Small projects? Old projects? New projects? Wikipedias only? And the corollary: Where is it *not* a problem? what are the characteristics of those projects where Problem X is not considered a significant problem? And finally: Will this actually fix the problem?
I'm not convinced this is the best way to spend about 3% of the total budget of the entire Wikimedia Foundation, especially when less expensive solutions, and ones that do not involve the WMF essentially taking over its biggest, most active community, have not really been tried. Sarah has made some interesting proposals in another thread that have the chance to develop some momentum if nurtured carefully. This...well, that's an awful lot of money.