Most KM practitioners insist on dealing with the organization in its entirety as the key abstraction for exploring and measuring KM initiatives. Why this obsession with bigness? It is either a fallacy of the highest order, or a convenient way of generating more revenue per consulting engagement. If we want to understand something as complex as KM then we need to focus on a more granular abstraction (i.e. the project team). Physicists seeking a better understanding of the universe decided to explore the smallest unit of matter they could find and were amazed by what they discovered.
Knowledge transfer always seems to happen most effectively within small communities sharing a common social context, whether among members of a high performance team inhabiting the same physical work environment, or a closely-knit virtual community working on similar problems. Social interaction, constant experimentation and feedback power the learning process.
If you want a return on your KM investment then you should focus on empowering these communities with the tools and resources that they deem to be KM enablers. Imposing a one size fits all solution is a complete waste of time and money. Only the insiders know what will work within their learning ecosystems. That said, KM czars certainly have an important consultative role to play. They can make suggestions related to emerging technologies and / or anecdotal evidence of what has worked elsewhere.
Once a community decides on an approach, financial resources and warm bodies are critical to the prototyping and implementation effort. It is OK to think global as long as customized solutions are driven locally, and by individuals that are the most likely to be impacted (i.e. teachers and students). Mission critical projects rarely get this kind of support. World-class knowledge workers always search for tools and techniques that will improve both personal and group productivity. They never lack ideas that would enhance both, if only they had the luxury of pursuing them.
Do not be conned into believing that you can engineer your way to KM nirvana. That is utter nonsense. First of all, most corporate cultures have a complete disregard for the humanities in the workplace and tend to treat all forms of labor as replaceable parts of the machinery. As if the automatons that you created are going to magically turn into renaissance men and women. There are many things that need to change before corporations begin to resemble institutions for higher learning.