My 2 cents
I have give the subject of academic wiki some thought in the last 6 months.
I mean a wiki that could replace the paper journal and its publisher. But
it would need a different collaboration model, In any sense that affects
thier income the Academia is quite conservative - almost backwards.
A researcher would not want to see his data sets vandalized or corrected
anonymously. Such a wiki would require both public and private zones
private - with complex setting to allow group based collaboration. The data
would be useless to anyone wanting to replicate experiment results without
documentation and access to the original analysis code. It would be
neccessary to have anonymous peer review. It would require a new bunch of
licenses - since many universities are interested in having patents for in
quickly and transparently sharing thier discoveries.
There are also social matters such as a prestidge and fashion and a great
resistance to change of publishing in your community. To get it going would
require an large community of influential academics who are not afraid to
chalange the large publishers and thier editors who are the current stake
holders
I would imagine developing such an acdemiwiki could be to developed quite
quickly once MW matures. I would need cloud based storage, a source control
system, advanced means of authentications. P2P capabilities. But almost all
of these things now are avaiable.