I don’t see how one “elicits beauty” from something that isn’t there.  Plain text doesn’t have enough power to encode even simple mainstream texts, which frequently include the use of italic, for example.  Yes, one can fake it, but then its not plain text anymore.

I’d like to see a format that at least allows unambiguous encoding of mainstream texts, capturing the author’s intent.  Yes, once again one can fake it using HTML, but HTML contains SO MANY other weaknesses in the other direction!

If we had an unambiguous encoding which captures authors intent, then it would be easy to go the other direction and “throw away” author’s intent when it doesn’t fit into plain jane text mode.