The story is usually cited as an early historical example of steganography. But a message in someone's memory is less obvious than a message on one's skull. A message in someone's memory doesn't seem more resistant to torture (a strong-willed slave might not disclose the message, while a strong-willed torturer would almost certainly discover the tattoo). It also doesn't seem likely that the message could be kept secret from the messenger (since the messenger could just ask a friend to read it off his skull).
In short, is there some information-theoretic advantage to this anecdotal method?