Depending on the content of the image, the compression ratio can vary a lot. The 26% figure is probably for "normal" photos. My images are mostly a few shades of black with a few white pixels (using a camera as a radiation detector) and I guess WebP is way better at compressing that than PNG.
I wanted to see if I could detect the radiation from a small sample of americium-241 that I pulled out of a smoke detector, so I put a Pi camera with no lens facing it and took exposures for a couple hours. After combining them and removing dead pixels I ended up with tons of tiny white specks where radiation had hit the camera sensor. I linked the final image below, and here's a timelapse video (compositing newer frames onto older frames to keep the radiation specks). video
You could argue it is a small advantage, but you cannot say it is not an advantage.
I would contend the advantage is actually quite significant. From time to time I am inside some building with poor cell service, causing images to take longer to load. 26% smaller images would load noticeably faster in these poor conditions
Some reports is that is saves space by half. that's a LOT of bandwidth. And for storage? It's cheaper now but as storage and bandwidth increases, so does the size of media. Higher def images and videos can now be easily and quickly shared so they are. Being able to decrease most images by half is an insanely huge advantage.