The article explains Tausk's views on this specific question from 1919 (schizophrenia, he said, involves "loss of ego-boundaries"), but surely others have looked at the question since then -- what have they found?
If schizophrenia has something to do with some self-referential part of the cognitive machinery being broken, it seems like this might be a very important clue to how the brain works algorithmically -- self-reference is very important part of the theory of computation (recursion theorem, halting problem, Godel, etc.), and understanding how the human brain handles it seems like it may be informative.