Because some water came out of the drive enclosure, I decided to let the drives sit overnight again. I've submerged electronics accidentally before and usually if you let them dry out before turning them on you are fine. The image below shows the drives as they were pulled from the burned NAS.
ioSafe N2 drives after burn and soak
While I waited for them to dry, I disassembled the rest of the NAS to get an idea of how it worked. The image below shows what I call the fireproof brick that the aluminum drive enclosure sits in.
What I call the ioSafe N2 firebrick
As I mentioned before, the aluminum drive enclosure then sits inside the brick (with a plastic liner in between). The enclosed aluminum drive enclosure acts as a heatsink for the drives and you can see the many short cooling fins. You can see an image of the aluminum drive enclosure sitting inside the fireproof brick below.
ioSafe N2 drive enclosure after burn and soak
As for the components of the NAS, they aren't designed to survive a fire. In the image below, you can see the wiring started to burn, as well as the charred remains of the motherboard. This sat on the bottom of the NAS enclosure with a thick galvanized steel plate separating it and the firebrick.
ioSafe N2 charred remains of component board
After a day had passed, I loaded the two recovered drives into the known-good NAS. I hit the power button on the NAS and after the N2 booted, I was able to check the contents of the volume and saw all of my data intact... amazing!
After a few minutes however, one drive started clicking badly and was automatically removed from the array, which you can see in the picture below.
Synology DSM showing degraded volume
Having two more good 1 TB drives, I started to disassemble the case to hot swap the clicking drive and repair the array, not knowing why the one drive was failing. When I got to the inner drive door there was already moisture accumulating on the door. In retrospect it appears that the water that got into the drive enclosure had not fully dried and the heat from running was sweating it out of the drive. Seeing water inside the drive enclosure I should have set the drives in rice for a few days to ensure they were fully dry.
The array repaired successfully and I left the inner and outer drive doors off in case the other recovered drive had moisture in it. The second drive continued to work after 2 days of sustained operation however. The lessons to be learned from this are two drives are better than one, make sure the recovered drives are completely dry and save the data somewhere else as quickly as you can just in case. If I were going to keep the NAS, I would have replaced the other original drive so that I removed any chance of latent failure from the torture test.
The gallery below has lots of images of the burning, the submerging, and the teardown.