The project, while interesting as an idea, was pretty much doomed from the beginning. My initial review was negative; Wikia Search was simply very far from a finished product. I've revisited the service after a couple of months, and though certain improvements were made, it was still just a promise of something that might or might not happen in the future. In the meantime, Jason Calacanis' Mahalo was doing pretty much the same thing with a lot more success.
Unfortunately, time seems to have run out for Wikia Search. As Jimmy Wales put it, "it was going to take at least another year to two before it's usable by the public, and we can't afford that right now." Long-term projects don't weather recession periods very well, and the average 10,000 unique users a month - which is what Wikia Search's numbers were according to Wales - simply wasn't enough. For comparison, other Wikia projects pull in around 30 million unique users a month.