I was blown away by this article in the NYT about Paul Otlet, his plans for cataloging all the world's books and his ideas for a system that easily connected all these documents, complete with details on their relationships and a way to query them. Of course, this sounds exactly like today's web, search engines and efforts to make all information "universally accessible and useful".
Sadly he ran into some problems:
1) The technologies he was using (index cards) were infinitely less scalable than what we have today
This just goes to show how ideas, no matter how grand, are rarely original. You need two things to make them a reality. First you need good execution (as VCs are fond of saying). Second, and just as important, you need good timing.