Hoi, This is a topic that deserves its own thread. Citations are problematic on so many levels it is hard to decide where to begin.
When the question is can we get a lot of citations in Wikidata, relevant citations, the answer is yes. It takes a lot of work and a lot has been done by the lovely people of DBpedia. They have data ready to link citations from Wikipedia to statements and it is fairly easy to gamify this. It means that it will be people who judge if a citation actually fits an existing statement.
When this is to be done with a high level of confidence, it helps when links and red links get the same treatment as interwiki links. They would all be associated with Wikidata. It will improve the quality of all Wikipedias measurably and it will also do a lot to improve disambiguation. The improvement may not be that much in the wiki links compared with the interwiki links but the red links will bring us the confidence what articles exist in other Wikipedias that have many red links in "your" project.
Citations are part of the quality issue. When Wikidata is to improve its quality there are several other approaches that will increase the confidence of the quality of Wikidata. When multiple sources agree on a statement, the statements where there is no agreement are the best targets for attention. With a relation between citations in both Wikipedia and Wikidata (the DBpedia link), it will be easy to signal to a Wikipedia for help with such issues. This may bring our communities more together as well.
Citations itself have a problem of their own. What do we do with debunked sources. Do we agree that a program of a conference is a source. How do we signal unease about sources when an author proves to be "problematic". What to do with the inherent conflict of interest that is pushed by people in an industry who want to include their data in Wikidata..
A fringe benefit of providing links and red links with Wikidata items is that it becomes easy to update the lists in use by "Women in Red". Many more of the subjects they want to write about will become available and, when an article is written it is easily removed from the lists. They already use this approach by using the Listeria lists.
When you read all this, you may ask yourself is this technically hard to do. It is not. Much of the software / functionality / data already exists in an embryonic way. BUT it works. There is only one question left: Are we ready to fix these quality issues?
Thanks, GerardM