More competition in the last mile would make a big difference, but its a very expensive undertaking likely to face predatory pricing once you enter the market, or even while you're building your network (which could scare off investors).  Wireless is less expensive than fiber, but has real limits, due to spectrum issues. Clearwire has begun building out wireless networks in some mid-sized markets, but its spectrum is constrained and it's actually been pretty upfront about blocking competitive VoIP providers.  Their long-term viability remains to be seen.

The best solution is for local government to get involved, either building networks on their own, or in some sort of partnership with private companies.  

Among the reasons for this are that a muni-network can be partly cost-justified by cost savings and efficiencies generated for local governments (e.g., remote video/data for police, fire, building inspectors, traffic mgmt, remote monitoring, etc.) and other "public" entities, like schools.  For example, in St. Cloud, FL, a recently launched govt-funded muni-WiFi service is offered free because the city believes it can pay for the network through such savings.

On top of these "direct" savings and efficiencies are less direct "public" benefits (e.g., economic growth, increased tourism, more jobs, better healthcare and education, less traffic congestion, gas expense and wasted time commuting, etc.).  

These are considered "externalities," meaning that they don't accrue directly to network users in a measurable way that can justify charging them fees for these benefits, but are more in the nature of benefits to society as a whole.  

That's one of the key roles of government--to invest in benefits that are "externalities."  That's because these benefits cannot be "internalized" by private profit-seeking companies through direct charges to users (e.g., the "market" can't effectively serve these public benefits).

So, the basic idea is that part of network costs are justified by government/public cost savings and efficiencies, plus all the "externality" public benefits resulting from a ubiquitous high-capacity neutral IP network.  That leaves only the remaining cost to be financed by the kind of commercial services targeted by private companies (i.e., the voice, video, data "triple-play").  Then the economics can start looking a lot more compelling than a privately-run profit-seeking "third network."

Some states have restrictions on muni-nets, which have typically been lobbied for by incumbents pipe-owners.  Both pending bills in Congress also address this issue, though in different ways.  

What's important in terms of legislation, is that neither state or federal laws put undue restrictions on muni-nets, which I'd suggest are the best hope for real competition and provision of a truly neutral high-capacity access network that addresses the nub of the NN issue.