I worked on a library that had to process PDFs and grabbed a bunch of PDFs from archive.org and other open sources for testing, and as I recall, about 2% of the

I worked on a library that had to process PDFs and grabbed a bunch of PDFs from archive.org and other open sources for testing, and as I recall, about 2% of them had at least one JPEG2000 image. So if you're in the PDF space it's probably still worth considering.