They’re only superficially similar
The two questions are only really similar is asking for build advice in their current versions. They’re different in ways the Stack cares about:
And if the closed question is suitable for the site now, it just hadn’t been long enough to get reopened.
It was correctly closed as Primarily Opinion-based
The original question literally asked for opinions and was correctly closed as such. One of the five votes was for it being unclear, and comments asked for clarification.
The other question was well-asked from the beginning.
The scope of the closed question is a drag on its reopen chances
The following is an analysis of the psychology of the reader (i.e., potential voters and potential answer-writers), not precise analysis of scope complexity. Few readers will do more than a rough intuitive estimate of how complex the scope is — which as we’ll see, is the point.
So now it’s having a bit of a hard time getting reopened due to, it seems to me, its larger area of possibilities for potential answer-writers to grapple with before coming up with a good answer, or even considering whether to answer. It’s asking for which archetype to pick for an unknown future up to 20th, which is 18 levels + feats or ASIs that answer-writers have to at least briefly account for for each archetype option (else someone is going to come along in comments and say “no, they shouldn’t pick Elephant Mesmerist, it’s a terrible choice because at level 17 it can only hypnotize two elephants, when Trapeez Artist can befriend three by 12th”), with optimised interactions with two other unknown-future PCs. The answers aren’t even being asked for a full build, just the best archetype, but they still need to think about this.
A full analysis might be necessary to answer the question, or it might not. It’s not required for an answer to be on-topic, but it might be needed for a “best” answer. And the thief could potentially multiclass at any level, exploding the possible full-level builds answer writers might consider before recommending just an archetype.
That doesn’t make it necessarily too broad — a lot of those possibilities are useless branches that can quickly be pruned — but analysis paralysis is real and this question asks for answer-writers to take on a lot of it before even deciding whether to answer. And same for voters deciding whether to reopen.
The linguist question by comparison is cut-and-dried: it is a given set of 19 levels with a wildcard 20th level. Its still a lot of work, but the scope is well-defined and not a matter of reader (or voter) judgement. Deciding whether it’s well-scoped is easy for voters. It had way, way fewer future unknowns for readers to analyse before deciding “I can answer this” or “this should stay open”. Its easy to judge.
And all that is stuff that answer writers probably don’t need to put in their answers, but maybe need to work out to make their answer best. It’s a lot of up-front work even before starting to write. Or voting.
The point, again, is not whether I think it should be open or closed. (Especially since you didn’t ask that; you asked why they’ve been treated differently.) The point is that analysis paralysis appears to be slowing its reopen votes.
There’s no one-size-fits-all help you can give
Every optimisation question is different. Useful advice depends on the unique circumstances — not just for what it’s trying to do, but as a question here on a Q&A site. Similar questions may be very different Stack-wise.
To help a held or closed charop question with advice, pay attention to the circumstances of its closing. Look at the edit history and comments. Leave a comment asking what’s wrong, if nobody has said. Bring it to meta asking what’s wrong with it and how it can be fixed, and assume good faith, rather than assume it has been closed wrongly.
After all, closing is normal here and even when it should be reopened, it’s just normal process in a vote-based system, not a flaw.