I entertain the notion that outside view might be a bad way of analyzing some situations, the post is a question on what this class might look like, and how do we know a situation belongs to such class?

The Outside View's Domain

inside view by definition has no evidence of even as little as lack of systemic bias behind it.

Not 'by definition'; if you justify using IV by noting that it's worked on this class of problems before, you're still using IV. Semantic quibbles aside, this really sounds to me like someone trying to believe something interpersonally justifiable (or more justifiable than their opponent), not be right.