For the completeness of links between MO questions, I posted an answer to a related question showing that Scott's formula (denoted (‡) in the above question) is realizable over $\mathcal{K}_2$, even with a computable realizer. So I guess it could be an answer to the present question.
↧
