Too many topics, too little time

Map of Science:
The Map of Science is a network map showing the relationships between 1.6 million scientific articles. The algorithm for the map was put together by Richard Klavans and Kevin Boyack. At the site you can click through to other maps showing analysis by geography, industry, institutions, and topics.
—-

Pretty cool, but does history guide policy in this way. does prior performance of scientific networks determine how we should encourage them in the future? I’m not really certain, but my opinion is that the data that comprises systems like this is at best representative of certain scientific misconceptions and at worst only related to individual subjects in a way that makes real systemic understanding impossible.