The strongest adherent of the scientific method admits and accepts that no scientific hypothesis is absolutely true. He does not feel threatened by the emergence of some bizarre theory, and is probably more effective in his refutations by allowing for the possibility of a new and perhaps unlikely hypothesis.
In defence of the poor scientist-in-real-life, he (or she) may not feel /threatened/ by bizarre theories, but he often feels /exhausted/ from constantly refuting them, and even /annoyed/ when called upon to refute (IHO) really stupid ones. But I think that our NPOV method can still deal with this.
In the cube example, we don't allow anything with no supporters (that falls under Wikipedia's ban on original research), and we don't feel the need to refute things with no arguments (it's enough to state the fact that the position is a fringe one). Then once the arguments for the cubical Earth are presented, we only have to lay out the counter-arguments once, there in the article. If the counter-counter-arguments etc get to be too long, then we simply spin things off into [[Cubical Earth]]. People that don't want to deal with this inane crackpot nonsense can rightly point its adherents to that article.
I don't either belive in NPOV. Never did and never will. Whatever happens Wikipedia will get the POV of the masses, a conglomerat of all authors working on the project. My task is therefore to add my pov on things to increase the sum of all recorded human knowledge that wikipedia is (coming to be). Over time most of the articles converge to something that becomes the POV of the masses - aka NPOV. Some wont, like the articles on drugs, because there seem to be a lot more potheads that write those than read them. Like the cubic earth example, writing "the earth is spherical" wouldnt be correct because 0,001% know it is cubic. Even worse to say that the holocaust has happened because 0,1% know it has. Or saying that God's existance hasn't been proven because 10% know it has.
Björn