Sorry I am so lame but it would help if you would state the problem and a proposed solution. Follow that up with details and dialog.
I think you spent a lot of words on meaningless prattle. It could be the Connectome is a promising path. I don’t see a case for it being the best path. You make a good case or at least a case for trying a different approach.
Yes, a great AI would be able to deal with the unknown. Humans can deal with the unknown but our record is not stellar. Only the great deal with it well and you propose that we need an AI that does at least that well.
Personally I never had to touch a hot stove to tell that it was a bad idea.
When the hardware is up to the task we will have AI that are more than adequate. Since it is so easy to transfer learning from one AI to another they will mature very fast. Start teaching a few today and do the mind merge in a decade.