(This is a reply to a blog entry where John Kerry solicited public feedback)
I'm glad you are interested in this issue.
I think the most important issues right now are network neutrality, and the 700 MHz spectrum auctions. There are other issues but not as hot as those two.
For users it is important to be able to use all available protocols and ports, without ISPs artificially crippling some of them. There are many ways this is being done: 1) in the worst case requiring use of a proxy server (so there is only access to the web, not to other services via SSH, telnet, NNTP, BitTorrent and so on); 2) prioritizing some types of traffic over others, so that the types the ISP does not "approve" will have worse performance; 3) blocking some types of traffic (BitTorrent, email service and VOIP being the usual suspects); and 4) nearly every ISP blocks multicast traffic. These tactics are all harmful, and have already balkanized the Internet almost since the beginning of commercial ISPs. Some of the blocking has been done (so they say) for benign reasons (to cut down on spam the ISP prevents the users from sending email directly - users must use the ISP's mail server or a web site like gmail). Even then it is a heavy-handed policy which does more harm than good. I want to see an Internet on which innovation is possible; it's not a single-purpose medium. Today far too many products must be designed around egregious limitations (such as in the worst case, all traffic must be shoehorned into HTTP on port 80, because the user might be stuck behind a proxy). I would propose going beyond the usual "network neutrality" proposals and require that every kind of traffic on every kind of port be possible, no blocking whatsoever. The issues of spam, worms and viruses, DOS attacks etc. can be handled in other ways. E.g. if the average consumer router blocks some services, that would be good for most users; but the advanced users who want to innovate should be free to do so, because the ISPs are not blocking anything.
The fact that analog TV channels are being freed of that payload comes at a good time for mobile wireless services. I am rooting for Google to win that auction and deploy some kind of mobile wireless Internet. I doubt there is much you can do to influence the outcome - it's between Google, their competitors and the FCC at this point. But I would have liked to see some requirements for the spectrum to be utilized effectively for digital public-utility wireless services, across the entire nation uniformly, rather than divvied up into little chunks for "public service" (exclusive to military, police and fire depts) and for the other bidders, if they win, to possibly just hold in reserve or use for who knows what, in such a way that we the people will not benefit. This is the best kind of spectrum there is - it goes through walls and can handle longer-distance coverage (blanket coverage of a city, for example). It's an historic moment that such a big chunk of this kind of spectrum is suddenly available, and the FCC's management of it seems a bit haphazard.
Another approach that needs to be tried someday is a more widespread free-for-all like what we currently have on the ISM bands. It's working well enough for WiFi, but there needs to be more spectrum allocation for that kind of unlicensed service. Spread-spectrum technology has made it possible for such services to mostly not interfere with each other, and when there is interference, to step around it. The old method of allocating narrow channels for each service is obsolete. Some management of bands is still necessary but the bands need to be wider and shared between more types of services using spread-spectrum technology wherever possible (again, not by fiat, but by permission, because it's the obvious best choice anyway - the market will choose to use SS if it can).
Finally there is the hot topic of broadband. I do not see a lot of room where the federal government can do much about that, except in ways that the playing field needs to be leveled (network neutrality, trust-busting and breaking up the grip of monopolies when they have behaved in anti-competitive ways). Seems to me the fundamental reason there are not more broadband networks must be right-of-way; it's not practical for too many utilities to share the same telephone poles, or to dig up the streets too many times to put cables underground. To the extent that the government is involved, it's mostly a local issue rather than a federal one; and probably on the cross-country scale there is already enough fiber to handle our future needs for a while.
Beyond that I don't understand why we are falling behind the rest of the world. Being a libertarian I don't believe in making the Internet into a government-run utility; I suspect some of the other countries have gotten temporarily ahead of the game by doing that - they saw what they wanted at this point in time, and just did it; but they may fall behind later because the possibilities for expansion and innovation are not as open - it's now a government monopoly. Openness and freedom in the past resulted in great innovation in this country. Let's make the freedom greater so the innovation can be even greater still.
I would also like to see some more restrictions lifted in the ham radio bands. The ban on "commercial" traffic has most hams afraid to access normal web sites using packet radio systems. The ban on obscenity is also archaic and has hams afraid of what they might accidentally run across on the web (while on 10 meter rag-chewing frequencies they are potty-mouthed in spite of the law). The ban on encryption stifles innovation. And there are too many regulations on the kind of digital protocol experiments that can be performed. I think that there is such an abundance of communication techniques nowadays that amateur radio should be all about experimenting, inventing the next new thing, rather than just plain old voice and morse code like what it's being used for now. There are enough ways to have a voice conversation across the globe, so that isn't as interesting as it once was, and nothing new is being done because too many of the possibilities are illegal. The licensing structure is in good form now - it's quite right to require demonstration of technical knowledge in order to get a license. After you've got it though, the usage should be wide-open.