I'm guessing the author suspects the more common first adaptation of JXL will be losslessly re-encoded JPEGs. In other words: the very subset you mention, right

I'm guessing the author suspects the more common first adaptation of JXL will be losslessly re-encoded JPEGs. In other words: the very subset you mention, right? Having said that, the author seems very open to suggestions and contributions - I suggested using canvas instead of a data URL-based PNG a week ago and within a day they had implemented a version of it. edit: the other reason might be having smaller image sizes in the cache.