There is one tip that really stood out and rang true for me because it has been something I have battled it at the last two major companies / projects I was working on.
It is number 6 under Key Aspects of Web 2.0 Strategies:
Incubators and pilots projects can help create initial environments for success with 2.0 efforts. Too much contact with the traditional support environment of an existing, primarily 1.0 organization makes it hard for 2.0 efforts to succeed; everything gets done in the traditional way instead of the new ways that are required. The traditional tools, processes, and skills just aren't there or are just too slow and burdened with unnecessary overhead. Creating dedicated incubators that are designed to use the strengths of the organization while being isolated from its weaknesses can help. Incubators are at risk of becoming too isolated however, and won't inform or change the greater organization unless care is taken to roll the lessons and capability back in.
I have worked at several organizations to push for keeping 2.0 initiatives external in some way to keep them shielded from the pull of the 1.0 energy in the organization. I get positive reception to the idea, but then when it comes to actually splitting the project or initiative off and make it happen I get resistance.
I think their needs to be a balance between keeping the project somewhat on its own, but have the benefits of the primary organization and its strength.
I don't know how to create the perfect balance for this to work, but willing to continue exploring.