I for one would certainly encourage further development of many different interfaces that support the collection of good edits from many editors across many different communities and I think "games" can be a part of that. While I understand the dangers of gamification done badly, done well, it can be a fantastic thing for rallying a community together. I would not advocate that wikidata call itself a game.. or even that it devote much if any resources to building game-like editing systems itself. I would however encourage the development of more applications like the wikidata game(s) that Magnus has created - but better. Luckily all of the things that I would suggest for improvements could be layered on top of the code he has provided.. For me 'better' would mean:
(1) Require some level of task redundancy as quality control. As the number of people playing these games grows, the increase in precision by adding in something simple like - "only make edit if 2/3 people agree" is well worth the loss of recall.
(2) make many games for many specific communities of interest and market them directly to them. You don't have to call them games either. You could just say "we need volunteers and here is a really easy way that you can help".
(3) If you are going to call them games.. (and that is a key IF), then make them more fun. There are lots of angles to take this - not just high score lists. This the principal double edged sword here that Lydia raised concern about above. If you make the game-element very appealing you have to be very careful that the game incentives are exactly aligned with what you want to encourage the community to do. If the game isn't quite lined up, you can get into trouble with people over optimizing to win the game while hurting the system the game was built to help.
http://zooniverse.org is an example of a collection of volunteer-tasks (for "citizen science") that could be gamified, but they have made the explicit decision not to do so because they are worried about incentive alignment - but also because they already have a community with more than 1 million registered users! While wikidata is off to a great start, there is a long way to go before we have the community needed to really achieve its potential. I support exploring many ways to get there, including well thought out games.
-Ben