http://code.google.com/speed/webp/gallery.html
WebP might be great, but how can it be a meaningful comparison --
1) take already compressed jpegs from wikipedia and re-compress them using WebP, and then tout the 39% file compression achieved? I can re-compress Jpeg with more Jpeg and also get 39% "more" compression.
2) as with all lossy compression, unless there is some measure of the loss, perceptual or otherwise, what are we really comparing?