Answer deleted by moderator because "not good practice"

Question

Today I noticed the following deleted answer: https://tex.stackexchange.com/a/561476/ (only visible for >10k users because it is deleted). Screenshot:

The question was asking about how to change a year value (2016) that is hardcoded in a class. I provided an answer myself in September 2019 about patching the code with etoolbox. The new answer was posted yesterday, so almost a year later, and suggests to edit the class file directly.

It was deleted by diamond moderator Stefan Kottwitz because it is "not good practice". Quote:

A class file should not be edited, there are other ways (such as redefinitions). As the way of doing is not good practice I remove the answer from public view ("deleted" but it stays here and can be improved). Thanks!

I was surprised to see this course of action by a moderator. Generally it is recommended that "bad" answers are downvoted, but not deleted, unless they are link-only or duplicate¹. A meta post by a different moderator makes this point explicitly in Correct flagging (or: bad answers are answers, too). That was in 2013, however more recently (2019) the point was made again in Should we really delete low quality answers?, i.e., the consensus seems to be that downvoting (to -1) should be the course of action for bad answers. Note that those posts seem to talk about actually wrong answers (i.e., answers that do not solve the problem), whereas the current answer is not wrong (it will work) but it is just considered to be bad practice.

So, the question is: do we as TeX.SE community agree with this policy of deleting wrong(ish) answers? If yes, what are the criteria for being "wrong enough" to delete an answer? If no, can this answer be undeleted, and can the moderators refrain from such deletions in the future (and regular users stop flagging wrong answers)?

Note that the linked answer is just an example. The answer is very short and lacks detail, so it may be better to delete it and convert it to a comment (although it is an actual answer, and therefore does not belong in comments according to 'official' Stack Exchange policy, but that is a different discussion). However, for this meta question I'm not just asking about this specific answer, I am interested in a general discussion on how to handle "bad practice" answers, the opinions on criteria for deletion, and the policy on using moderator privileges to make content rating decisions on an on-topic and non-abusive answer.

¹ These criteria apply to actual answers, i.e., posts that attempt to answer the question. Other types of posts are deleted as well, such as follow-up questions posted as answers, "Thank you" posts, spam, etc. This meta question is not about those types of non-answers.

Answer 1

For a more complete picture: that answer has been flagged two times:

• as being "not an answer"
• for having "very low quality"

So for me the completed question becomes: should a moderator delete badly shaped answers with bad content (see barbara's comment there to know why) after users complained?

You linked two good existing meta questions. Let's see what is said there:

• Correct flagging (or: bad answers are answers, too): The actual request there by our fellow moderator was: please don't flag such a case.

• Should we really delete low quality answers? The suggestion in the answer was to skip the review in such a case. And the highest voted comment preferred to "recommend deletion and leave a comment". (that's what I chose here too)

But, what should preferably not happen according to the meta suggestions above, happened:

• user review queue action was done
• flags were posted to tell that there's an issue

Once that is in the world, a moderator had to take a decision (such as keep it, delete it, or convert it) that may be a good choice for the site. Note, in 2019 we got 4,154 answers flagged and 15,114 comments flagged. We work on fellow users flags the best we can. And as here, we are open to discuss even a single flagged and deleted answer - as you see I take much time for that. If the case is not totally clear and agreed, we simply can undo the deletion, as I will do here, you see. At the end it's for us users.

So at the time we got the flags here, I decided to provide the helpful message above to the answerer regarding the issue of editing class files, in a light way and thanked them. I did not educate them on the not good style (missing punctuation, not capitalizing LaTeX) as it's not so important and it's a new user. Downvotes wouldn't be helpful either in being friendly. Converting to a comment would not be a good option because that bad advice stays and cannot be downvoted then to signal the issue.

That answer does not really match quality requirements to an answer on StackExchange. What shall we do with that issue in the flag queue?

Back to the one-liner answer by Ardashir Mohammadzadeh that lead us here:

Open the latex class ieeeaccess, find the 2016 and change it to 2020

So here we have this several times flagged answer with "dangerous" advice, because editing class files can lead to issues, as barbara wrote there:

There are two things wrong with this answer. First is a problem caused by the creator of the class file -- unless the article is actually accepted and published by the journal, identifying a purported volume and issue is unethical. The publication's editorial staff should be notified about the problem. Regarding changing a class file, even though it is to be used only by whoever makes the change, if the file escapes into the wild, it can cause serious confusion. Don't do it

I hope that explains the reasons for me to decide to delete it, with adding friendly information to the poster. And again, no problem with undeleting it on request like here. And that's what meta is about: talk, solve, improve.