The problem of one package being popular or not has been the discussion
in the background of the list for some time. I remember the speech
about psfrag, natbib, etc. But a little question: Are we the users?
We are usually the programmers. Even a math-oriented person like Robin
Fairbains is now a CTAN maintainer!
I think preparation of some question forms for distribution between LaTeX
users will have a great outcome. We can know which parts are used more,
and which parts less. This is somehow like taking a vote. So we can know
which package they use and love, which packages the use and hate,
which package they haven't heard of, which feature they like to see
in the future, etc. etc.
I know that some of us recommend them the use of some packages by making them
available on distributions (ranging from emTeX to TeXlive), but let's
forget that effect now.
P.S.: There has been always a problem in undemocratic societies (like my
home country): there have always been some people who think they know what
others need, and/or like; but they are usually wrong.