Duplicates: Should we close the new question if it's better than the old one?

Question

This has been on my mind for a while: if a new question is similar to an old one in meaning, but asked with far more detail and focus, should it count as a duplicate or not?
If not, should the old question (clearly not duplicate at the creation time) be marked as duplicated, left as is, or addressed otherwise?

A recent case in point:
New question - Why do propellers have less theoretical range (i.e. range ignoring impatient passengers wanting to go fast) than jets, despite greater efficiency?
It has been marked as a duplicate for:
Old question - Why don't airliners have turboprop engines instead of jet engines?.

The old question has a -2 rating and got closed. It was closed for duplicating a separate question (and not being very good), but let's leave that out of the context for now, it's just an example.
Gravedigging and editing the old question to match the new one in this case would amount to its complete replacement, and essentially result in misattribution.

What is your opinion on what our course of action should be in such cases?

I don't like dupes, but I don't like closing good questions in favor of worse ones either.

Answer 1

Maybe we shouldn't, but who or what decides what is better?

I share your judgement regarding the answers, but they need to be found.

The wording of the old question is quite specific and to the point. The new one is less helpful when searching for relevant content. When you hit the old question, just follow the links to the better answers.

The answers to the new questions are also problematic, being very specific to the (also very specific) question and (as so often) the better answer is not the accepted one.

I think you have a valid point but chose a poor example. In general, I would prefer to add links to newer questions to older answers, so the best content will have the most links.

Answer 2

I've reviewed some sites that deal with this issue on a regular basis, although with wildly different kinds of content.

In there, new content that duplicates old content can either be killed as a dupe, if it offers nothing new, or override old content as a trump, if it's broadly better. If both have competing redeeming qualities, both are kept.

There's also a broad notion of accessibility levels: some content is useless to advanced users, some is inaccessible to laymen, some is both.

Per previous discussions here, questions and answers are to be judged by today's standards, so being first does not establish immunity.

Based on these practices, I'd propose a following set of guidelines:

• Pre-existing questions should be evaluated together with their answers. If a question looks good, but has never attracted any quality answers, it can't be considered a good question+answer combination.
• New questions that don't raise any issue that hasn't been fully answered before, should be closed as duplicates. Phrasing improvements should be incorporated into the old question as an edit, if possible.
• New questions that are likely to lead to better answers, should be allowed to remain. Once they attract answers that completely cover the old question, it's fine to close the old question as a duplicate.
• Questions of a similar nature, but vastly different technical depth, can coexist.

Airframe-specific versions of the same question are a separate case from this, but should probably be evaluated with the same policy unless they address airframe-specific issues.

This is meant to be a basis for per-point discussion or alternate suggestions, of course.