This seems wrong on its face: it conflates intelligence and consciousness (hence the various premises about brains being computable). You don't need a machine to be conscious for it to be intelligent, or superintelligent.
Both your argument and the argument that you're criticizing make assumptions on what intelligence is and what are the properties that define human-level intelligence. How can you know that consciousness is not a prerequisite for human-level intelligence if you cannot define human-level intelligence?