What Federico is saying is really important in my opinion.
Wikisource can have a good impulse offering a huge set of different languages while other digital libraries are offering the usual and most used languages.
The best would be to approach the problem step by step and working with languages having weak communities where the aggregation in a single hub can offer them a real opportunity and keeping other biggest communities still indipendent to solve their problems.
Even if the community decides to look for an aggregation, what would be the next step? Merging all together with an un-manageable change?
I think that the good solution is to improve the central project and to introduce the best solution to have several languages and to contextualize the page, afterwards other linguistic projects can decide to look for an aggregation (may be in connection with the decrease of size of their communities) or to keep their independence.
Kind regards