That 12 MB floppy is cool! It looks like they adopted the hard shell and spring-loaded shutter from 3.5" disks, which is a good idea, given how sensitive a disk with that kind of TPI would be to dust and fingerprints.
The sticker information is particularly interesting. It reports itself as 333 TPI, which is less than the 384 TPI of a 6.6 MB disk. That forced me to figure out how a 12 MB capacity was achieved.
The 5.25" floppy disk's 6:5 ratio of TPI to cylinders results in 277.5 cylinders. With 277 cylinders and a 1000 kbps transfer rate, you'd get an unformatted capacity of 13,850,000 B / 13,525 KiB / 13.21 MiB at 300 RPM, or 11,541,667 B / 11,271 KiB / 11.01 MiB at 360 RPM. Neither of those seem like capacities you would market as "12 MB". Split the difference, and 330 RPM (5.5 RPS) yields 12,590,909 B / 12,296 KiB / 12.01 MiB. Could it be?
The next hint is the 78 sectors. 78 256-byte sectors with 277 cylinders would give a formatted capacity of 11,062,272 B / 10,803 KiB / 10.55 MiB). With the 360 RPM unformatted capacity, there would only be 4.2% overhead, which is unrealistic. IBM standardized on 26.3% overhead when they chose 9 sectors for their 5.25" and 3.5" DD disks, and 15 and 18 sectors for their 5.25" and 3.5" HD disks, respectively. Kaypro and DEC went with 10 sectors, translating to 18.1% overhead. Overhead is 20.1% if we go with the 300 RPM option and 12.1% at 330 RPM. 12.1% is lower than Microsoft's 21-sector 3.5" DMF disks, which had 14% overhead, but it wouldn't be unprecedented for Drivetec. Their #3218R floppy for the 320 Super minifloppy was formatted with 18 512-byte sectors and an overhead of only 11.5%.
Most likely, these disks had 277 or 278 cylinders, and spun at 300 RPM with a 1000 kbps transfer rate, and a formatted capacity of 10.55 or 10.59 MiB; roughly 11 MB. There's also the tantalizing possibility that Kodak dared to reduce the gaps and synchronization bits enough to cram 10.55 MiB of data onto a disk with an unformatted capacity of 12.01 MB by spinning at 330 RPM @ 1000 kbps -- making it also notable for having a non-integer RPS -- or 660 RPM (11 RPS) @ 2000 kbps -- making it also notable for having double the data rate of a 2.88 MB DSED 3.5" floppy disk.
Everything about this is weird. Even with perpendicular recording, this is probably pushing the limits of what's achievable using servo tracking. It's starting to push into floptical territory.
