I have tons of images and tons stored in fractal compression formats which solved the space and quality issues years ago and now I can't open any of them unless i use Windows 98. Also isnt this new jpg just doing what jpeg2000 promised that nobody really bothers with either? Or just use a jpeg stripper to remove all exif data and save a ton of space without the image data being touched which works great until you need to know the date or camera settings of a photo. speaking from experience. For Google it would be useful for web accelerators that don't perceptively effect web images and faster ad delivery to your apps and games. I would rather see a more robust lossless compression archival format with ongoing future support that the images can be retrieved. At least jpeg is fairly universal and good for its task plus its robust for data recovery where even partial data loss doesnt destroy the whole image unlike most image formats. Then there's PNG, its too early to dredge out my brain
I do like some of the projects Google comes up with, not so much this one is all.