Great post! This reminded me of Nick Bostrom's slightly scary book on Superintelligence.
We are quite a long way from being able to upload a brain of any organism and that is probably a good thing. The major problem appears to be the difficulty in the creation of a neurocomputational model. The technological challenges in scanning and processing are there, but the technology we have should be enough for extremely simple brains. The OpenWorm project, whose goal is to simulate the brain of an organism that has only 302 neurons, shows us exactly how complicated this process can be.
Precisely. Of course, the connectome of C. elegans is entirely finished and has been simulated for seconds as a proof of concept. The motor function of the process that is the only part being simulated as far as I can tell is at the moment at 30% at best. A stable simulation seems super distant.