The numerical representation of the content is not what's in question. It's the user-perceived content. Let's say you take the contents of a book, add an extra

The numerical representation of the content is not what's in question. It's the user-perceived content. Let's say you take the contents of a book, add an extra space between some words, then compress/hash the new contents. The numerical representation could be completely different, but the user-perceived content is the same. Sorry to get all legally lawful on over your ass, but the law of the land (and common sense) dictates that the new contents are still the works of the original author even though the numerical representation is completely different.