Most content disputes are associated with *behaviour* that oversteps one or more Wikipedia policies, and while there is no arbitration or mediation process for content, there are numerous dispute resolution systems that do focus on behaviour.
A problematic editor who is "wrong", is also probably violating the policies on "no original research", "cite your sources", and "neutral point of view". It also usually works out that a person who persists in adding invalid material to articles in violation of these policies will also violate the policies against "personal attacks" and on "civility". This isn't always the case, and unfortunately it often happens that the dispute becomes so heated that nobody maintains the moral high ground on these points.
Nevertheless, I think most content disputes can be treated as behavioural issues and resolved on the basis of existing policies without anybody needing to decide the actual content issue. The [[Capitalism]] article is not a test case for content disputes because the problem editor has violated a number of Wikipedia policies, such as "cite your sources".
A real test case would be one where a person was politely and civilly putting a kooky point of view into an article, while cheerfully admitting that it was only one of the valid points of view that must be presented under NPOV, and happily citing sources -- all kooky as well. As long as it was not original research, and there were some sources for it, however laughable, such a dispute could not be resolved without a decision on the merits. At present, the only mechanism for that is weight of numbers: the side with the most editors wins.