I agree with bringing this up, not because I don't think we should host them but because for a while they've always felt like a strange borderline item. Might as well get a ruling on them out of the way.
The positives of these questions are simple to enumerate.
- Objective - It either is, or it isn't. No opinion technically changes this.
- On-Topic - It's a question to identify a game.
- Is A Question - Big round of applause for these to actually be questions. They also rarely can be judged by the NARQ Marks (This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form.)
- Is An Answerable Question - We have a decent enough track record of answering these. At least 50%, if I'm not mistaken. Which, well, isn't stellar.
- Is A Solvable Question - The correct answer to these are conclusive to the question.
But these are basically our bare minimum expectations we're hitting. So rather than know what the positives are, it's enumerating the negatives that directs their worthiness to remain on the site.
Localized - I don't find this a strong argument. Nigh any tech-support question we get has this same problem, wherein they basically pray that one of us had the same problem. Likewise, some of our more esoteric questions have very low searchability and likelihood that someone else will need the same answer. At least, until they read the question which leads to them being interested.
Low Usefulness - I have argued the importance of being useful as a metric for good questions. However, I stated in just the same point that all this does is emphasize what is worthy of upvotes. Less-than-useful questions, such as esoteric ones, can still be fit for the site.
When it comes to personalized questions, there's a difference between when I ask for something that still has a measure of external curiosity it can inspire ("How do I make this difficult but ultimately unnecessary climb in Bunny Must Die?") and something which is really localized ("Given my party setup, how can I defeat the Behemoth King without Quickenings?"). In both cases there is low searchability, but the former is one that people who do come across the question will learn something, while the latter is just a bit too "Help me win my game". I find that identify-this-game tends to lean towards the former.
"Ease" - It can be stated that these are "easy" questions, in part due to them being based more on passive knowledge than active knowledge. But, in that our goal is to provide answers, I don't see this as a problem. With the exception of your anomalous example (it has the "A-For-Effort" vote accrual), these questions don't actually get a lot of votes.
Bad Trending - I can't predict the future, but there is the possibly valid concern one might see with regards to the volume being encouraged. There are only 84 such questions right now, which isn't too much. While both game-recs and identify-this-game have uncountable permutations, the latter has a smaller infinity (think of rational numbers versus integers, for example). To me, this is only a concern if they become some kind of farm like List of X questions. I think we can err on the side of allowance for these because they don't generate enough attention to be a real problem.
Passive Knowledge - Now, here is the trouble point, and the one that does concern me since it is one of the major fall points for game recs. The majority of our questions are based on what I call "active knowledge" - people provide excellent answers either because they deliberately learned it in the past or they performed research. In essence, it refers to true subject matter expertise. People are more inspired to answer these.
There is subject matter expertise in knowing about games themselves than knowing how to play them. The problem is that our site does not exercise this knowledge, which is why we perform so poorly at repository building. No one is inspired to do all that much research for identifications and recommendations, and instead people rely primarily on their passively acquired knowledge. If you passively learned of the game (such as by playing it in the past), then you provide answers, but if you didn't learn of the game you just skip the question.
Identify-this-game has a lot higher success rate than game recommendations ever had. But its reliance on passive knowledge does worry me, as it makes them far less likely to get answered (which, also, is a similar problem with tech support!). It also may make it feel unwieldy to some users, since people get correct answers for "just knowing". I'm not certain how much weight this carries, and if it's enough to bar them. But it feels like the primary issue.
Not A Problem - The other real issue can probably be that these feel less like real problems. They're outcasts even among trivia like what were characters named, because at least other people are going to search for that. It's also rarely a pressing issue for the author, too, as it is far more likely someone is comfortable to wait and see if it'll ever get answered. This can lead to stagnation in combination with the passive knowledge point: no one cares enough to answer, and no one cares enough to dredge it back up.
This is my analysis of the situation. They are real questions, even if they aren't problems so to speak. The issues of searchability are not too different from a variety of other questions we cater to, including the kinds of questions which define our existence (the obscurity of a game is irrelevant to its fitting on the site). However, they have mediocre performance, are based more on passive knowledge than active knowledge, and can stagnate really badly. I think they have a point, but I can understand if it is agreed that they don't have enough of a point to stick around.