Is the question "which features should be added to package XY" adequate for TeX.SX?

Question

The question title actually states what I would like to know. Let me give some background and describe the use-case.

As author of pgfplots, I feel honored that this site attracts both power users and beginners, and I feel honored that lots of people appear to find my package useful.

I am also aware that a lot of questions about pgfplots turn out to be feature requests (compare Questions which are feature requests). Some have cool work-arounds by power-users, some provide best-practises - and in most cases, the suggestion "file feature request" appears (with good reason).

Since pgfplots is a spare-time project, the priorization of incoming feature requests has three major criteria: (1) the efford to implement it (cheap features are added immediately), (2) the "fun factor" for me as implementor (I like to implement cool features even if they are of limited use) and (3) the "votes" of end-users.

I can access and control (1) and (2). Concerning (3), I have difficulties to get a suitable overview.

This leads to my question: is it acceptable for a Q&A site like tex.sx to host questions of sorts "Which feature do you miss most in pgfplots?". The votes on such answers would probably give a relatively good view, and since tex.sx has a high visibility, it would also ensure that it is a representative. But I can understand that it is ... well, not the common question for a knowledge-base platform.

Answer 1

After a feature has been implemented, the corresponding answer/suggestion would be obsolete. To not keep a mix of implemented features and open requests, the answer should be removed then. This would remove reputation of course. No problem, it could start as a community wiki question.

Regarding fitting the site - would the content of the question and the answers be helpful for any other reader than to the OP? I doubt so, though the result may do on the long sight. Possibly the question could be treated as "too localized", because answers only (directly) help the package developer, and they would be valid for a certain time or certain package versions.

Regarding votes, early answers may get many more votes than late answers.

Another idea: you could implement such a survey on your pgfplots website, where it can be visible and maintained for years, and announce it via channels like your mailing list, blogs, Usenet and web sites.

Anyway, it would be more beneficial than a lot of questions which perfectly match the site rules, so I won't have a problem with it. I just thought about the SE view and having a hundred more people doing the same, even others than the authors or maintainers.