I find myself wondering if gamedev.stackexchange is overly cautious of
There's a lot of subjectivity in game development, even within the technical domains like programming. Often times that subjectivity is either discussion-based (which has no place on an SE site) or highly-contextual. The latter could be appropriate here and could be a strong contender for "good subjective," but that context needs to be specified or the question falls back into the extremely broad, discussion-oriented kind of query that's a better fit for a forum.
to the detriment of the site
It's one think to not like that we close a lot of subjective questions. It's another to assert that doing so is damaging the site in some way. I'm not saying you're wrong, but I'm also not seeing any evidence for such a claim.
I'm not saying high vote-count makes a question good for the site, but
it's a helpful filter to find some questions where one might dispute
their being closed.
That's good, because vote count is definitely not a metric. Consider also that many of those questions are very, very old relative to the site's total lifespan. That both means they've had much more time to accumulate votes and also that they come from an era of the site when it's focused was being defined.
Further, the number of concrete times somebody has actually attempted to bring up any of those questions for re-opening? If you include your post as such an endorsement, I can probably count those occurrences on two hands instead of one. But generally it doesn't happen that often. There aren't any pending re-open votes on any of those questions and they are very rarely referenced in meta topics like this.
Good games to earn your wings is, while broad on the surface,
something that a lot of experienced developers talk about when giving
advice to those who want to learn to make games. The answers there,
while broad, are an asset to this site.
You can make a pretty strong argument for most closed questions that the answers are an asset to the site. That's not what the site exists for, though. It is not trying to be all things to all people, not trying to be the one-stop resource for all game development questions. It's trying to be a very specific thing, focusing on specific and mostly-objective (ideally canonical) questions with similarly-described answers. There are plenty of other excellent sites for discussion and broader "theory crafting," for lack of a better term, within games, and we aren't trying to replace those. Indeed, we often try to direct people there when their needs are pretty obviously at odds with what this site provides.
How to prevent 'too awesome to use' syndrome is a good game-design
question, and good answers have been given to it. It was closed for
being primarily opinion-based...
I don't remember closing that, to be honest, but re-reading it now I probably still would. I think the issue with the question is, as I've noted before, a lack of appropriate contextual scope. It's presented as a question about a mechanical aspect of game design without any sort of scoping through a genre, through the specification of additional extra mechanics, and so on. I think that makes it too broad (so I concede that 'opinion-based' is not what I would use to close it today). But that was two years ago.
Does the existence of helpful and yet more subjective stack exchanges
+ the existence of helpful and yet closed game design questions on this exchange perhaps suggest we should open the doors to less
objectively answerable questions?
Potentially. But how would you do that? What is your proposal for "opening the doors," exactly? I don't mean to be harsh, or to pick on you specifically, but but when you want to change a policy you should probably have some kind of idea how you want that change to look like in practice.
I look at these questions and I look for things that give the question a lot of context. Ones that put the question in a framework defined by the game being developed. Questions that provide enough of such context, I think they make "good subjective" questions. Questions that don't, I think they make "bad subjective" questions.