Extracting separate parts for multi-voice staff

Please see my slightly intentionally cryptic responses in this other thread for more on this topic: