I think there may be something else at play here.
First things first, in the spirit of @Oli's answer to the question @muru linked in his comment, I agree that these old questions should not be "closed": they were on-topic at the time of posting; no need to update their status for that reason.
As time passes, a lot of content becomes obsolete: we should expect that and embrace that, as I have written about it earlier in this answer.
However, I agree that we do need something: something that could be a feature-request.
That would be a new flag-variant. This new flag could be labelled as such:
Prevent the Community bot from upcycling this question to the front page.
Because that indeed seems to be an issue.
There were earlier meta conversations about whether the Community bot should upcycle old questions at all, and the consensus seemed to suggest that yes, it should, because many of the old questions still contain relevant / useful information. (I think this may be the specific thread that I have seen earlier.)
At the same time however, there are those old posts that are entirely irrelevant today, e.g. because they are asking about some subsystem that got entirely abandoned, and is not part of Ubuntu since a long time.
These latter posts are pretty irritating on the front page, I agree, and can trigger one's OCD: "something is not right with this. This is a kind of a problem that could be sorted out with a little piece of code: so it should be". This is the very drive that makes us write software. I believe ignoring this feeling would not be right.
That's how the new flag variant could help: it would allow keeping the habit of upcycling old questions to the front page for the useful parts in them, but would allow us to vote out old content that proves to have grown entirely irrelevant.
This flag could work in a comparable fashion as the close votes: as in, at least n
number of people should cast it for it to take effect. In turn, it could work without needing manual effort from mods.
It would be interesting to consider how many other stacks could benefit from such a feature. If there would be more such stacks, we could actually hope that SE takes on this feature request: it does not seem to need a huge effort — it could be achieved (re-)using/expanding already existing features.