It loses any data that is not structural in the neurons' physical shape - whose importance is not currently known. We can presume that electrical signals can be rebooted, but can chemical ones? Will the brain fail as badly as a drunkard or someone who drinks twenty espressos, if shorn of its chemical context?
This is particularly plausible because the brain is full of low-level feedback loops controlling endocrine stuff - I could fully expect them to go completely bugfuck if their sensor inputs suddenly read "0.0".
To give an example here: "gonadotropin-releasing hormone analogue" drugs are used to block secretion of sex hormones. They were originally designed to increase them. GNRH is the "on switch" signal, the drugs mimic it. And indeed they do initially increase hormones - then the brain's regulatory feedback slams on the brakes, all the way to zero. Nobody knew that particular feedback was there before they poked it with a drug.
This tech may make testing the above much easier though.
Freeze, slice, stain, and microscope can check chemicals in the way this cannot.
This will end up being important if the details of these systems are important for encoding the differences between humans rather than just important for having a properly functioning brain. If it's just important for proper functioning, we can just figure it out once and then assume that the brain you're reconstructing has these systems in good working order. If it's important for encoding the differences between humans then we'd have to preserve these systems to preserve "you".
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?