This feels like the wrong end to optimize. Zip is plenty of fast, especially when it comes to a few hundred pages of a comic. Meanwhile the image decoding can take a while when you want to have a quick thumbnail overview showing all those hundred pages at once. No comic/ebook software I have ever touched as managed to match the responsiveness of an actual book where you can flip through those hundreds of pages in a second with zero loading time, despite it being somewhat trivial to implement when you generate the necessary thumbnail/image-pyramid data first.

A multi-resolution image format would make more sense than optimizing the archive format. There would also be room for additional features like multi-language support, searchable text, … that the current "jpg in a zip" doesn't handle (though one might end up reinventing DJVU here).