First Amendment Issues: Free Speech
Very briefly, this right is supposed to protect the highest levels of free speech, such as Whistle Blowing, followed shortly thereafter by Parody, Satire, and other commentaries. A little farther down are intended protections for simple original works, possibly with some risque or provocative language.
That is the intent. However, over the last 15 years, a number of problems have been growing. Let's start with some of the easiest.
A safe place to start is Reckless Endangerment. It is fairly well established that yelling "Fire" when there isn't one, purely for the entertainment value, isn't a quality use of free speech. In these essays, we will be getting into the subjective tone as well as the objective statement of the laws, and there are several cases when the subjective tone seems to be veering in a dangerous direction. However, this doesn't seem to be one of them.
In a *fictional work*, having a character perform this stunt, would be mildly offensive, but otherwise the audience would know they are not in danger, and wait to see the Author's intended denouement. In a real setting, it hits the Cry Wolf problem, so that very real dangers involved in stampeded arise solely for someone's entertainment. From what I know of the social landscape, there isn't much opposition here.
Slander & Libel
Some of the tougher categories begin to arise with Slander and Libel. Both are Defamation. (Slander is more transitory and verbal, Libel is more permanent and written.) The basic tests here are that these speeches are Malicious and False. Essentially, if you make up junk that threatens someone, you're headed toward these categories.
On the adult level, these are seemingly fair exceptions to free speech - you can't just say "________ is secretly gay" just to gain political points. Nor can you say "The supervisors in 3rd world factories of ____ Company rape their workers." Excepting actual news stories, these statements are False and Malicious. So far, so good.
Troubles arise pretty quick. Any schoolchild will relate many instances of these, most of which are never punished. On Internet sites, "Trolls" regularly post examples of these kinds of statements, displaying an animalistic low-cunning level of communication, partially trying to rile up innocent naive net citizens, and on the other hand, trying to rile up moderators and administrators into taking drastic action. This is where the more serious actions arise.
One of the "safer" ways to handle this is trough something called User-Moderation. In general, users moderate a post up or down, so that for example, if a post is moderated down to -1 and your reading level is set to 0, you don't see the worst of the junk. Sure, there are problems with biased moderators with agendas pushing quality posts down. That's why some sites have a Meta-Moderation feature, which means that a second person gets to look at the post, and basically agree or countermand the original moderation. This is intended for selected quality moderators to review maliciously down-modded posts, and restore them back to something visible, and penalize the original moderator for being abusive. There are still flaws, but it is an important approach.
A Simpler Way
Much simpler, and also fraught with more direct risks of judgement, is that the site admin team moderates posts. The resulting quality of discussion depends on them. With good "Mods", you get a decent stream of comments. But with that centralized power comes the dark risk of Censorship. The Mod team behind a big newspaper might politically decide to delete a damaging but true comment. That's where the slippery slopes begin.