/my current layman understanding
The current most plausible path for uploads seems to me to be brain scanning, which currently operates basically by a sort of plastination process, then slicing the brain into extremely thin slices, imaging each slice at very high resolution, and then assembling it in a computer into something meaningful.
The imaging part is currently in pretty good shape, even if you need to mess around with toxic metals to get something you can use an electron microscope on, and the slicing part has been automated too; one of the remaining bottlenecks is, well, how to process the brain to get it into a slicable state in the first place. This is a question that plastination quality bears upon.
So the only step in the process that needs to be done now instead of post-singularity is the one we're worst at?
Comments? If superior brain preservation can be demonstrated under a 5nm-resolution 3D scan, plastination wins over vitrification hands-down. Is Robin missing anything here, or is this indeed as important as he says?