Why measure scientists and their work?

in I'm interested in using graph theory to develop tools to improve science (whether I manage to do this is a moot point, I believe that someone will, and that such tools will be useful). One of the questions raised by any effort to introduce new systems that may be used to compare and contrast the work of scientists is whether such systems can have any net benefit when weighed against the potential pressure that they may place on working scientists. A large amount of discussion has taken place in the Citations in Science forum. Other places where there is a discussion about the kinds of tools that are being considered is the International Repositories Infrastructure wiki, and again over at Nature Network Martin Fenner is asking a few questions about author identifiers.

Themes that frequently bubble up are:

There are other themes too, and I may amend this post as those themes come to me.

posterity vs a living wage

inbound vs outbound effects on the scientist.