Like, literally, the idea these guys were developing, that almost came to market, was the idea of storing data on an optical disc not as a binary 1 or 0, but as 8 different levels of mark to non-mark on the disc surface. This allowed CDs up to 2GB, DVD's up to 7.1 or 10GB, and single-layer blu-rays up to 60GB. Some prototype drives were made, but the tech never made it to shelves. And I wonder if we were robbed something cool or spared something awful.
Optical drives and players can be picky enough about discs just having to determine if it's a 1 or a 0, I don't know about trusting them to be so extremely precise, even if they were specific lasers and optics made for that purpose...
Just seems like an idea that made more sense in, like, Laserdisc, a varying wavelength of reflectivity works for analog video that can take dropouts like a champ, but if your 80GB BD-R DL MLR is slightly smudged then the video Goes Away because H.264 doesn't like it when the data goes bad for half a frame because the laser thought that 6 was a 7.
As a result looks like a lot of the extra bits gained doing this had to be devoted to error correction anyways, and the drive firmware had to be way more complex to do sophisticated power management and "adaptive equalization" and stuff to even work at all, which probably should've been the sign it was a bad idea.