Academic papers are hidden from the public. Here's some direct action.

90% of the 8 billion dollar world-wide market for academic research is paid for by library subscriptions. There is a serious problem here, indeed. Some excellent facts and recommendations were posted in this report. It was written by the Association of American Universities (61 leading research universities) in January, 2010 and it was initiated by the U.S. House’s Science and Technology Committee in conjunction with the White House’s Office of Science and Technology Policy (OSTP): http://www.aau.edu/WorkArea/DownloadAsset.aspx?id=10044

Are there any known instances of a lecture recording being used against an instructor unfairly?

I doubt doing things that could be construed as "out of line, incorrect, inappropriate, whatever" really improve your lectures in a significant and sustainable way. In any case, there are better ways to be interesting and entertaining [1].

I think everyone understand that instructors aren't concerned about copyrights. They're concerned about having to improve the quality of their lectures, or have a permanent record of lectures they're not really proud of.

[1] Richard Buckland, UNSW: http://www.youtube.com/watch?v=2bLCjMA0YlE

I don't know about that. Thinking about the four universities I've got experience with, the only one that didn't have someone checking IDs at the door of the library was the small one that probably couldn't afford it.

Not really. This is about being able to access the specific papers written; it is not about being able to read an aggregate of the research without necessarily being able to look at the original source (which is what Wikipedia is).

(EDIT: grammar)

Indeed, Wikipedia's particular style won't work for science.

The shame of modern science is that what might work is so simple to build, because it is already built. Research papers already form a web: They all have references to prior research papers, which in turn have references of their own. Moreover, the community of researchers and their students would quickly and happily turn all those links into clickable ones -- if only paywalls didn't prevent them. [1]

It's no coincidence that even the simplest of web pages is already perfectly suited for scientific research: Tim Berners-Lee built HTML and HTTP for that purpose. But now science lags behind, say, LOLcats or the Twilight books when it comes to online discussion, and I assert that it's not merely because science is esoteric [2]: It's because scientific publishing is suffocated by rent-seekers.

---

[1] Actually, my understanding is that cross-journal hyperlinks are slowly evolving within academia: Once you're inside the university library's firewall you can make actual working hyperlinks that are only, say, 65% less efficient than the average Wikipedia link. This is great for those who are currently inside the ivory tower. Not so great for me and the other 99.9% of humanity.

[2] What I wonder is the extent to which science remains esoteric because it has always required so much training and money to read the literature that the literature has never needed to evolve actual readability. Why bother learning to explain your work to a popular audience when your writing will only ever appear in journals that are only visible to your colleagues?

Historically, we've relied on great science journalism to make up the difference. (Gods bless you, Isaac Asimov.) But great science journalism, like all journalism, is rare and getting rarer.