Using the Max Lossy settings will generally get you similar savings to what the new Pagespeed recommendations will tell you.
However, I wrote a post over in the Pagespeed forums explaining the flaw in trying to get a perfect score, as it relates to image optimization, and I’ll try to give a shorter version here.
Basically, Google’s tool is attempting to determine what quality level your images were saved at, and comparing it with what they believe to be the optimal quality level. The problem with that method is that it generally only succeeds when the tool doing the detection is the same as the tool used in compression. Thus, if you use EWWW on your images, and Google is using jpegoptim, they will never get an accurate reading, and they might recommend further savings because the image quality is still so high. Other users have seen the same behavior with Photoshop, saving at quality 60 gives them better quality than the Pagespeed image, and higher savings, but Pagespeed then tells them they didn’t do enough compression, even though they already beat the original recommendation.
On the flip side, several folks have reported great success with TinyJPG in the Pagespeed forums, and since that is what EWWW uses at max lossy, it should generally satisfy the Pagespeed metrics.
So, can I guarantee that Pagespeed will stop complaining about your images? No, but that isn’t the point of Pagespeed. The point is to make your images smaller, and your pages load faster, and I know EWWW’s max lossy setting will nail that.
