@Alex So, the essential point of disagreement between us is whether to consider B a duplicate of A in a case where answers to B would be valid answers to A but would not be reasonable to expect in an ideally complete answer to A. In that case, you would call B a duplicate of A, and I would not.

Given that it's unreasonable to expect that B's question will get answered under A in the normal course of events, even given ideal answerer participation, doesn't your policy violate your Goal #3 in that case?

@IsaacMoses Sounds like a good summary. But there will also be situations where my approach turns out to be more lenient. In a case where you could reasonably expect a complete answer to A to address B but there still might be an answer to B that wouldn't address A, you would say it's a duplicate while I would say that it's not a duplicate.

@Alex Could you construct an example of such a case?

@IsaacMoses It would certainly make it harder to achieve Goal #3, but it doesn't rule out Goal #3 because there are still methods to encourage an answer to be posted to A (comment, bounty, chat, etc.).

@IsaacMoses Let me think about it.

@Alex Why would it be desirable to require unreasonable means to get some questions answered?

@IsaacMoses It's not desirable. But it's necessary to uphold Goal #2.

To make it more concrete, if we've already got "May one eat rabbits?" with a really great answer that addresses all the major opinions and every step of mesora from Sefer Vayikra to Meniyas Achilas Arneves Kehilchoso, and now someone has a question about whether medical exigencies could permit rabbit-eating, why is it better for Mi Yodeya to try to get that person's question answered as a left-field answer to the original question rather than as a new post?

@Alex Goal #2, taken that strongly, would suggest that if a Halachic principle is addressed in answers to one question, it would be bad for that Halachic principle to show up again in answers to another question. I don't think that's right.

@IsaacMoses In that case, perhaps the second question should be posted about medical exigencies in general, and not specifically about rabbits. Otherwise, we might end up with a site where every single question has a parallel asking about medical exigencies in that particular situation. I'd think it better to have one (or a few) question(s) about medical exigencies in general where any cases can be addressed.

@Alex What if there's reason to suspect that medical exigencies interact with rabbit, in particular, differently than with other classes of issur? (Not that there is, but I could come up with something more realistic.)

@IsaacMoses Ideally, we would not want the same halachic principle to be used differently in different cases (assuming that there's not actually a halachic reason why it would be different). But we can't stop it in a case of different questions, because then we would be violating Goal #3.

@Alex I disagree. I think it's inevitable and desirable for individual Halachic principles to show up in different combinations throughout our body of literature (as they do throughout any body of Halachic literature except the very most minutely codified).

@IsaacMoses Then that would be a specific answer to the general medical exigency question, and we're back to my original position. But if there truly was something specific to rabbits, it might already be reasonable for a general answer to the general question of medical exigencies to address it.

@IsaacMoses I agree with that. What I meant was that they shouldn't show up differently just because two different users disagree about the principle.

E.g. we wouldn't want an answer to one question saying "XYZ because safek d'oraisa l'chumra" and an answer to another question saying "XYZ because safek d'oraisa l'kula".

That's a case where one user is simply wrong about a halachic principle. A reader might get confused reading those answers, because they are in contradiction to each other, but we can't help that (other than to post comments explaining why one user is wrong, and downvote accordingly).

@Alex I'm not sure I understand what you're getting at with Goal #2. Gotta go, but will revisit this later.

We can't close all questions that could feature safek d'oraisa in the answer based on an answer to another question that lays out the rule of safek d'oraisa l'chumra, because there might be an answer to a new question that doesn't deal with safek d'oraisa at all.

@IsaacMoses I'll try to explain it in forthcoming comments; feel free to continue the discussion at your convenience.

Goal #2 is: The same question should not have different answers in different places.

Let's take the tefillin on Chol Hamoed example.

If we have two different questions asking about tefillin on Chol Hamoed, it's possible that one question will contain an answer saying that we do don tefillin on Chol Hamoed, while the other question might have an answer saying that we don't wear tefillin on Chol Hamoed.

This is undesirable because it leaves open the possibility of a reader only seeing one question, and thus only seeing one answer. Such readers might be missing out on pertinent information.

If one of the questions is closed as a duplicate, then there will only be one place to post an answer; thus, if conflicting answers are posted they will be in the same place. Readers will then have access to all the information.

Of course, it is possible that there could be only one question open and the answer might still be lacking some information. But that is simply a case of a non-comprehensive (or wrong) answer, and there's nothing we can do about it except post (or encourage others to post) the missing/correct information.

The above is all in a vacuum (e.g. where the two questions are exactly the same).

But most of the time the two questions are not exactly the same, and therefore Goal #3 becomes a factor as well. As I mentioned in my answer (and I don't know if you agree with me on this point), in my opinion Goal #3 is the most important. Since this is fundamentally a site for asking and answering questions we need to ensure that every question (that is otherwise on-topic, not primarily opinion based, not too broad, etc.) can receive an answer.

So if it comes down to a conflict between Goal #3 and Goal #2, my opinion is to sacrifice Goal #2.

However, as alluded to in one of my above comments, if Goal #2 can be upheld without violating Goal #3 even though it makes it harder to achieve Goal #3, and the alternative is sacrificing Goal #2 entirely, then Goal #2 should be upheld.

Now, to take an example, let's apply this to the rabbit case.

If we close B (rabbits on Tuesdays) as a duplicate of A (rabbits with no qualifiers) Goal #3 is made harder, but it can still be upheld.

If we leave B open we are now subject to the possibility that an answer to B will contradict the answer to A, which violates Goal #2. (Goal #1 would probably also be violated because a good answer to B would probably reproduce much of the answer to A, in order to explain that Tuesdays are an exception rather than the general rule.)

Therefore, upholding Goal #2 (and Goal #1) outweighs making Goal #3 easier, and B gets closed.

However, to take another example, let's say A is "What beracha is made on an apple?" and B is "What beracha is made on an orange?".

Both questions should remain open (in lieu of a third question "What beracha is made on fruits?") because it is possible to answer the question about oranges without addressing apples, and vice versa.

Now in this case, both Goal #2 and Goal # 1 are violated, because it is possible that the answer to apples will be reproduced to oranges (e.g. "the beracha for fruits in general is ha'etz"), and it is possible that the questions might get contradictory answers (e.g. an answer to A says "the beracha for fruits in general is ha'etz" but an answer to B says "the beracha for fruits in general is ha'adama").

But since Goal #3 is more important, we sacrifice Goal #1 and Goal #2 in order to uphold Goal #3.

@IsaacMoses Here's an example (it might not be the best example, but I was more pointing out a theoretical difference rather than one that is likely to occur):

We have the following question:

My answer there says that there are 13,550 seifim in Shulchan Aruch.

Theoretically I could have had my answer as just one sentence: "There are 13,550 seifim in Shulchan Aruch."

If I wanted to avoid all the downvotes I would probably get on such an answer I might add: "I know this because I counted them."

However, someone might assume that it would be reasonable for a complete answer to this question to provide more details.

@IsaacMoses Kehilchosah

Indeed, my answer provides additional information, most notably, the breakdown between each chelek.

Thus, you might reasonably expect to come away from an answer to this question knowing that Orach Chaim is the biggest chelek.

Now what if someone else wants to know about the size order of the different chalakim. He can post a question such as "Which chelek of Shulchan Aruch has the most seifim?"

Someone following your reasoning (even if you personally wouldn't take it this far) might say that such a question should be closed as a duplicate, because it is unlikely that this wouldn't be answered by an answer to the existing question.

But following my line of reasoning this would not be a duplicate, because it is possible to address the size of the different chalakim without ever mentioning the number of seifim in Shulchan Aruch — or even the number of seifim in any of the chalakim.

Someone could post an answer saying that Yoreh Deiah is the biggest chelek.

How does he know? Because for every seif he added a block to a tower, and the tower for Yoreh Deiah is taller than the tower for Orach Chaim.

This can't be posted as an answer to the original question since it doesn't tell us the actual numbers.

In fact, now that I'm thinking about this, we could make the example even better by changing the first question from "How many seifim are there in Shulchan Aruch?" to "How many seifim are there in each chelek of Shulchan Aruch?"

Now it's basically guaranteed that an answer to the first question will address the second question.

But I would probably still say that the second question is not a duplicate because there is a potential answer to the second question that would be invalid as an answer to the first question.