In the context of "directories" I prefer to think of it as taking a concept to a logical extreme.
If Google is the gold standard of information discovery, retrieval, indexing and access - with the latter defined by a page with nothing more than the iconic search box - then what's to hold everyone else to the standard of looking not Google and acting not Google?
Heaven forbid, whilst being not Google, those that are not Google fail to serve to Google the information architecture, taxonomies, titles, metas, keyword density, etc. that Google says is so important to the WWW. To everyone else, that is, but Google - so far as "websites" go.
Who knows. Maybe the next evolutionary step in the development of the increasingly botted WWW, heavily mined by non-human "visitors", will be the emergence of query dependent websites - en masse.
What if a directory didn't look like a directory? What if, instead, it looked like a directory fashioned on-the-fly, like Google's SERPs? Entirely possible.
Interesting to think how this might play out with future implementations of existing forms of software. Say a forum where there was no index, just a query box or a comment box, with a backend that filters or sorts and "files" comments. Say, "me too" type comments are sorted and filed into an enormouse "Me Too" thread - so the accumulated value of the added comment eventually adds up to something?
Does a directory need a "directory front end"? I guess so, since structure can be an aid to discovery. I guess a directory that is driven by a well designed database would also be one pleasing to Google, but what if databases or DBMs evolve towards more "loose" or "fuzzy" models, something akin to emulating a human brain/mind?
Maybe taxonomies, site architecture, navigation, presentation layers, frontends and all the rest is overrated and a burden in an information economy. Maybe Google will be a driving force leading to that eventuality.
Just a thought to play around with. Maybe it really is true, at least in the end, that it's the content that matters?
Back to the real work of grinding out keywords and taxonomies and link navigation structures and . . . Bleh. ;)
[edited by: Webwork at 5:07 pm (utc) on Oct. 24, 2008]