Simplify, Objectify and Stabilize Policies and Guidelines
That hasn't really been my experience with the project. I've found IAR to come into play in enough circumstances. No one full out ignores verifiability, but you do ignore it in the short run as you're working on bringing articles up to quality. And when new users run afoul of policy, it's not because they added a source that doesn't meet the nuanced "36mph while wearing a hat" rule. It's because they didn't add any sources at all.
As long as verifiability is understood properly (everything can be checked in a reliable source) it should never be ignored. If you do misunderstand verifiability as to mean that listing references is all important, then please ignore it.
And yes, some users, and consensuses, do ignore verifiability. And yes, you can get into trouble for any kind of reason at all, depending on where you are. You can get in trouble for failure to violate NOR, failure to violate V, or NPoV, for removing a false reference, or an empty link, for correcting "2+2=5" to "2+2=4". Anything at all.
There are really weird consensuses in Wikipedia. Anybody can edit and anybody can form any consensus about anything. There are no limits to weirdness, as long as a bunch of users get together and agree on something. - Brya 05:04, 24 May 2011 (UTC)
»You can get in trouble for ... correcting "2+2=5"«: I'd like to see that! Reference or diff, please.
I have seen things like that, but the point is that in practice a consensus can be anything whatsoever. Not just through the groupthink “A mode of thinking that people engage in when they are deeply involved in a cohesive ingroup, when the members' strivings for unanimity override their motivation to realistically appraise alternative courses of action” but also by the nature of the political process; politics is governed by interest, not by content, so if a group feels that a weird decision serves their interest, it will make that weird decision.
A consensus-system limits the contents that are possible within Wikipedia; just as "2+2=5" is rare, as it will usually be felt to be outside the consensus, so will anything of real quality be rare: the consensus will stop it. - Brya 06:24, 25 May 2011 (UTC)
If by your first paragraph you mean that, as a whole, people tend to do what they think serves their interest, then I agree with that. That's a truism that does not only apply to "the political process". Your second paragraph is not only cynical, it's also impractical. It only could potentially be meaningful if you had a clear definition of "quality" that is distinct from "useful for a substantial group of people". Clearly, "2+2=5" fails that by a long shot, and it will therefore never survive in Wikipedia. SebastianHelm 02:15, 26 May 2011 (UTC)
I have long ceased to be surprised at what survives in Wikipedia, for really long periods. Putting in "2+2=5" will not survive, but it needs only be rephrased a little and it stands a good chance of surviving.
I understand that there are many who do not know what "quality" is and who do not want to know, and in Wikipedia there are all too often conditions where it cannot exist. Still, it is a pity. - Brya 05:56, 26 May 2011 (UTC)
Actually mathematicians operate with additive operations wherein 2 (+) 2 = 5, see e.g. en:Abelian group.