Matt Brown wrote:
What I've personally considered is that under certain circumstances, while original research cannot remain in an article, original research can IMO justify leaving something OUT of an article. In other words, I think that if someone can prove, or show to a good level of certainty, that something is NOT true, no matter how well sourced, it should be left out - even if that proof is done with non-published sources.
This is precisely the sort of thing our original research policy is meant to prohibit. We're here to summarize what has been reported, not to mount independent investigations to determine "the truth" on our own. In physics, for example, we report what is widely written in physics books and journals, and also significant minority views. If in your own unpublished research you've determined that some of this is incorrect, that's great, but doesn't justify leaving the information out of a Wikipedia article unless and until you publish your new findings in a physics journal, at which point we can then cite your view as a contrary one.
In other words, you should be making your attempts to "prove, or show to a good level of certainty, that something is NOT true" somewhere other than Wikipedia, and then we can cite them.
-Mark