Hey guys,
I think for a long time semantic projects have focused on getting data out but haven't incorporated the 'voice of the consumer'; yet, if you think of data quality as 'suitable for customer requirements' instead of 'process conforms to specification,' this is the first step.
I'm looking forward to getting an early look of the data you publish and otherwise advocate for the point of view of people who want to use Wikidata data.
I'd also like to advocate for "documentation driven development" here, because my experience with
http://code.google.com/p/basekb-tools/
is that it helps a lot to
(1) document the behavior of the system by providing specific query examples, (2) construct test sets for everything in the documentation (3) If you're not 100% proud of the story told in your documentation feed this back into the product.
I think a little attention towards producing "readerly" output in conjunction with a "writerly" interface for Wikipedians could be key to make Wikidata one of the foundations of the Linked Data Galaxy.