I think it's unique to JPEG due to its simplicity. JPEG encodes blocks independently, which makes manipulation of block data easy (also enables progressive rend

I think it's unique to JPEG due to its simplicity. JPEG encodes blocks independently, which makes manipulation of block data easy (also enables progressive rendering). All later DCT-based formats try to get more compression by predicting future blocks from previous ones, which makes all blocks potentially interdependent, and distortions from such quick recompression could be amplified exponentially.