[Feature] Optional removal of blank pages from the archive file (originals untouched)

Description Problem Duplex scanners frequently produce blank back-sides. These blank pages bloat the archive, add noise to the document, and make multi-page documents harder to read. Paperless already passes clean, deskew and rotate_pages through to OCRmyPDF, but there is no way to drop blank pages — OCRmyPDF itself has not taken this on: the recurring requests ocrmypdf/OCRmyPDF#12 and ocrmypdf/OCRmyPDF#1032 are still open years later, and ocrmypdf/OCRmyPDF#1121 was closed as not planned. Existing community workarounds run as pre-consume scripts. They share one hard drawback: they mutate the source file, so a falsely-detected blank page is lost forever. There is no safety net. Proposal Add an opt-in, default-off setting (working name PAPERLESS_OCR_REMOVE_BLANK_PAGES) that removes detected blank pages only from the generated archive PDF, after the OCRmyPDF run. The original is never modified — it stays the full source of truth and can always regenerate the archi…