Some years ago we designed a 5-year "experimental" project in NYC—with $50 million in Annenberg funds, to explore a large-scale experiment with the above in mind. It involved an institute at Columbia headed by Linda Darling-Hamond, Michelle Fine at CUNY and other research backup, about 130 schools with 50,000 students, organized into 15 networks. It gave schools direct access to their full budgets and a great deal of freedom from union, city and state mandates in return for developing new forms of accountability. It managed to get the support of the then-chancellor, mayor, teachers' union and state commissioner. Unfortunately, as we were about to "go", both the chancellor and the commissioner departed and their replacements said "no way."
It's an oddly distorted version of this idea that emerged 10 years later under Bloomberg and Klein. It gutted what we believed was the essence of the plan: that it was voluntary, small scale (the size of the average American city), invited networks to develop self-designed plans, and had the support of some of the best independent research institutions in town to track different aspects of the work as it played out over time. We hoped that the work would help us find answers suitable to the various audiences involved. We were genuinely curious and thought it quite likely that we would end up with some shared agreement about "what works" and many different answers as well!
This theme plays out over and over and over, with infinite variations.