Sientific data for future use and re-use

I am interested in data curation and warehousing. More specifically, so that such preserved data can be reused and combined with other data to find new solutions to problems and new insights in to what makes the world tick.
Much has been written about this. Recently my colleagues and I (members of the Alliance for Information Science & Technology Innovation) published a white paper on this.
My question is, how and how fast might this set of innovations diffuse and become almost common practice:
- understanding the value of current research data at Universities and setting policies that will capture such data and preserve it for reuse?
- using such data in collaborative, distributed future science projects on a regular basis?
- capturing all the data from research projects as you go along and handing it off at the end of / or at significant points in the research process?
- academic libraries will facilitate this process?

I wonder if complexity theory and/or Everett Rogers' Diffusion of Innovation theory could inform how this will emerge in future? There are obviously precursors to such a system that function well already but in large scale data repositories such as the genome project.
Do you know of good examples of mid-sized research projects that archived their data and made it available in an open access system?
Johann