Am I nuts? Looking for outcomes of using NG library sites vs "old" library sites

[ I apologize if this is in a FAQ of some sort =) ]

Am I nuts? I'm looking for outcomes of using NG library sites vs "old" library
sites.

I say "NG library sites" to mean the "one-stop shop" the library puts up
on the Internet--with info about the lib's services, the "stuff" it holds,
guides to "using it", and ways to get to information pros/librarians, etc.
based on ideas or discussions which have been touched on here.

I guess I'd like some help in finding literature to:

1) Find results for a survey taken by "gurus" in "the field" about what kind
of stuff would be in this one-stop shop

2) Find results of some comparative tests with a NG protoype library site
vs. the "old" library site/opac/databases. (To find out if it really was
better and how it was better). And, perhaps, at the root of the matter, did
users use this NG library site more than the alternatives (e.g., Google).

I see people implementing Scriblio, Endeca, etc. but what are the measurable
outcomes?

I guess I would like things like:

* Is the time spent in finding a library book in the OPAC reduced when:

    -... there is faceted search vs. no facets?
    -... there is user tagging vs. no user tagging?
    -... there is "other users who checked out this book also checked out
____" vs. when there is none?

* Is the "give-up" ratio the same when either or all of these features in
place? Is the repeat usage (or satisfaction?) greater when either or all
of these features in place?

* Is there more user loyalty, or satisfaction (or "using Google less") when
users are exposed to these "NG" tools for.. say, 60 days?

Any pointers would be welcome =)