It seems that many questions I considered duplicates aren't viewed as such by a decisive amount of users, therefore I'll need to adjust my response to these questions1 and I would like to ask for some guidelines on how to answer them.
Let's consider for example possible questions that could appear under the Harry Potter tag:
A) How many Horcruxes had Voldemort made?
B) Describe the Horcruxes made by wizards/witches in Europe in the XX century.
a fair amount of questions like these could be created, all of which probably could be regarded as different, standalone ones, while they still could be answered by the same, often already existing answer. It is not a problem, as it is often repeated here on meta:
answers don't make questions into duplicates.
Ideally, an answer to questions A) and B) would contain data such as the name of the Horcrux creator along with the description and names of all the known Horcruxes appearing in chronological order, supported by canon references. Now, we have an answer here by Slytherincess, which perfectly answers both questions with maybe some minor adjustments.
Another example would be the recent question Why does Turanga Leela and her family put their surname first? which probably would be adequately answered by gnovice's answer especially after their latest edit, with simply emphasizing the 'we don't know about the existence of any canon origins' part as a header.
So what would be the recommended action, when a user2 posts a question if a good answer on the subject already exists on SFF?
Should we direct OP to the answers of existing, similar questions and try to answer the new question only if OP states that the existing ones are unsatisfactory or perhaps we should attempt to answer the question right away, without an attempt to enlighten OP about our current resources which may provide an answer to their inquiry?
If we decide that an answer is required to OP's near-dupe question should we utilize our stack (which is full of quality answers many of which kept up to date), that is, base our answer on existing answer(s) that fit3 by perhaps asking the author of the original answer to repost their answer4 to the new question with slight modifications or maybe we should quote or reword the old answer with attribution / link 5 , adding only parts which are lacking in the old answer? Would that be recommended or perhaps we should 'reinvent the wheel' and ignore altogether the original answer and write a similar or maybe even identical answer based on independent research?
What would be the best approach?
1. Until now I've simply VTC'd them as dupes
2. especially if it's a new user, see Rand's answer
3. an analogy would be maybe not duplicating code
4. If they're still an active user, naturally
5. links would be also useful for maintenance, (i.e. adding newly released info on the subject to all the relevant near-dupe questions for consistency and clarity)
'an answer to this other question could mention what you're looking for' is even worse justification
the author of the answer writes in a comment that the reason it's not mentioned because a canonical answer is not available (yet). Sounds reasonable, as in SFF it's the norm rather than an exception to support an answer with canon refs. but if you're skeptical of that happening and therefore in favor of having another question on the subject to ensure the addition of a canonical answer, shouldn't we open even more questions to increase our chances?