How would Satoshi prove himself then?
If he had the foresight to know he might want to prove his identity at some point in the future, he could have embedded the hash of a message into an early block (using the public keys of coinbase transactions). The message could say something like "Satoshi Nakamoto is <real name>. <salt>". In order to prove his identity, he'd just have to reveal the contents of the message (he'd have the make the salt something he could memorize so that it wouldn't be at risk of being destroyed by a fire). Ideally he'd encode this message hash in multiple early blocks (using a different salt in each message, so that the hash is different each time).
I think anything like the "knowledge of the content of private emails" is flawed. Satoshi's email account was hacked, so most of his "private" emails are no longer private (supposedly he didn't use PGP for most of his private correspondence).
If I were Satoshi I totally would've committed it to memory. I have more random characters than that memorized for passwords to various accounts.
Or, even easier, have a smaller simple password, and then be able to reconstruct them using an algorithm that your remember, e.g. SHA256(SHA256(SHA256("correct") + "horse battery") + "staple").
[0] http://organofcorti.blogspot.com/2014/08/167-satoshis-hashra...