Storing blobs outside the database

I had a PostgreSQL database with loads of scanned documents, as a document bytea column in the table scans, with hundreds of thousands of documents that was large and inconvenient to backup. Also, there was a high rate of duplication, not major but 5-10%. I wanted to store those documents outside of the database so they could be backed-up via incremental tar, and reduce the size of the pg_dump database backup. I came up with a solution I want to share below, using plperlu. Any comments or further optimisation ideas will be appreciated!