My first impression was that the combination of lossy compression and transparency would give WebP better traction (because before designers would have to choose between JPEG and PNG for those features). But lossy compression is mainly useful for photos, and transparency is mainly useful for graphic art. How often would a designer want to apply lossy compression to graphic art? How much would would lossy compression compress graphic art anyway?

Now I'm thinking that the combination of lossy compression and transparency will mainly be useful for adding transparency to photos (as demoed in the article).

From the article:

With today’s updates, WebP now offers a lossless mode as as well as support for transparency... According to Google’s test, the new lossless mode offers a 45% reduction in size when starting with PNGs found on the web, and a 28% reduction in size compared to PNGs that are re-compressed with pngcrush and pngout (both of which are popular tools for optimizing PNG files).