Google Fiber is revolutionary with the medium of its infrastructure: fiber all the way to the customer premise and the enormous headroom it offers to accommodate future bandwidth growth and new, high bandwidth services. This will allow Google Fiber to leap past the big incumbent national telephone and cable companies bogged down by their existing investment in wire cable infrastructure designed for a pre-Internet era. As Marshall McLuhan put it, the medium is the message. And the medium is fiber to the premise.
Google Fiber also has a revolutionary business model that lessens the pressure to get customer premises to subscribe in order to make the network economics pencil out. It allows customers to sign up for a low cost, multi-year, flat rate connection designed to cover the cost of connecting the premise. The big telcos and cablecos, by comparison, typically charge many thousands of dollars to connect a premise lacking access to a connection, with cablecos charging about $65,000 per mile.
Where Google Fiber's business model is not revolutionary is that like the big legacy incumbents, it is a closed "walled garden" that seeks to own the customer rather than an open access network that sells access to customer premises on a wholesale basis to those wanting to market services over it. This imposes major marketing costs to acquire subscribers one at at time, limiting Google Fiber to those areas where it can get a good return for its marketing and infrastructure investment.
Critics of this business model such as Michael Elling (featured in this video) contend it degrades the value of the network by limiting its ability to scale, invoking Metcalfe's Law -- that implies a network is only as valuable as the number of subscribers on it. Since fewer subscribers can connect to a limited, closed network, it becomes less lucrative in the larger scheme for those at the core providing Internet delivered services such as Netflix, Amazon and ironically, Google itself.