Comment by mr__y

This would come with a cost of longer compression times - either multiple attempts with random shuffling or pre-compression file ordering optimization process. For resources that are compressed once and then distributed and decompressed multiple times this would be quite interesting solution