What is our stand to edits that significantly change the question, with potential conflict to the original intent (read: change the core question), especially if the author is believed to not come back, ever?
I'll agree with Anne Daunted's answer that we shouldn't do this. It's not worth the effort. We could just as easily write a new question. I'm writing this answer, because in addition to that I have some further arguments.
I'm going to say we should, in general, keep to the 2 guidelines you posted:
- Edits should respect the original author's intent (no guessing)
- Edits should not invalidate existing answers
I've read this quite an interesting answer on Workplace meta. What I like here is the 'When should we edit' part:
- If the question is new, make a comment first, and make a quick edit to get the asker to participate
- If the question is several hours old and still open, only edit if it looks like it will be closed (it has no good answers, downvotes, close votes, etc.)
- If the question is already closed with no answers, have at it (anything is better than a no-answer closed question)
The highlights are mine. I think these parts are especially important to take into account when we want to think about editing questions as significant as the one on Workplace was.
That said, for the Workplace question, what I'm seeing is that the question was originally there (although this may be influenced by reading the edits). What I'm seeing there is that 'it just needed to be worded better'. Also:
- The most serious editing was done by a moderator. This is significant. Personally, I'm not confident enough to go pulling off that sort of 'tricks', and I don't think it's my place on SE to do such things even if I had.
- The editing was taken to meta, to discuss if it was a good thing to do. That means even the moderator might have had doubts about whether this was okay to do the edit/if the edit was good enough or not, in that specific case.
- All 7 edits were done within 1 hour and 10 minutes of posting the question: The question was posted on 15:39, and the last edit made at 16:48. I don't think your example question was edited because 'the author was believed not to come back, ever'.
- According to the
meta about that question, it was significantly edited 'to avoid it from being deleted'. That's a different thing than editing because something is on-hold but the user seems to have abandoned their question. I think the fact that a moderator did the editing is important here: Moderators can delete questions, so this moderator must have seen something that made them know this question was not just close-worthy, but delete-worthy.
An example of a question on this site that was heavily edited by other users: How do I tactfully avoid interacting with a person I meet on the way to work every day, at a subway station?
- The first edit invalidated this now-deleted answer. The gist of this answer was that if OP was so afraid of interacting with random people, she should seek help. Since we're discussing to close questions that need professional help, an edit that invalidates this answer might be, in my opinion, very dangerous. We've talked about the dangers of giving advice to somebody that seems to need serious professional help in that meta question. Although we can't always be sure that something needs professional help (an OP isn't necessarily always honest) in my opinion it doesn't help to edit out the parts that are suggesting it does need that. That's just a way of circumventing a serious problem.
- This question wasn't edited soon after it was posted, so it could be saved. Instead, the 22 edits are sprawled over 3 days.
- This question was put on hold, but again there was no indication that it would be deleted before the OP had a chance of editing it.
- This was not 1 big edit to save the question. Most questions or problems were addressed with another edit when they were raised (and some of those edits weren't done by the OP), mainly dismissing things as not a problem by stating things like "Please address that aspect rather than my possible motivations." and "I have a right to avoid him if I want to, whatever the reason."
- My last questions in the comments were never answered, with a comment by OP or another edit by the user doing 11 of the 22 edits. I think this is due to the fact that the user editing the question couldn't/wouldn't answer these questions, and the OP never reacted either. Editing stopped after that, and the question is now 'abandoned' by both OP and the editing user.
In my opinion, the editing, due to how it was done, didn't improve the question, there still isn't a very clear question there. Part of this might also be that the comments, edit history, and OP's reaction to frame-challenge answers seem to imply that the OP seemed to want to be validated more than ask a question about how to handle the situation.
So, on to the example from your question, where nothing has been done: How do I let someone I'm romantically interested in know I think we wouldn't work out because I am not straight?
- It's currently on-hold. But in my opinion (I'm not a moderator, I don't know the guidelines here) there's no danger of imminent deletion. We're patiently waiting until OP edits their question.
- OP seems to have abandoned it (no answers to questions in comments were ever given, no editing from their side). That's sad, but not our business. If we're in a similar situation, and the answers that are already there aren't helping us, we're free to write our own 'better worded' question. The comments as to what should be improved are still there, a new question could take these into account.
- If we were to edit it, it would not be a quick edit to get the asker to participate, this question has proved it was abandoned.
- The question is on-hold, but does have answers. So 'anything is better than a no-answer closed question' isn't applicable here either.
And a very recent question where things were done right: How to open up to someone about loving them, when you don't know if they love you back?
- We asked OP questions, close-voted immediately and asked them not to worry about the close-voting.
- OP provided info. We edited in the info to make this into a good question in 8 edits within 30 minutes-1 hour.
- No guessing was involved.
- No existing answers were invalidated (there weren't any).
- Only one edit was made by OP themselves, which makes this a good example of questions edited by the community and how community editing can keep an OP involved.
So, based on what I've found in regard to editing guidelines, and with this example of editing gone wrong in mind, I wholeheartedly agree with the NO from Anne Daunted. And we shouldn't go editing your example question.