I can understand your concern, but the only time we tweet anything is when you click a button asking us to do so; it's very hard to imagine how we might introduce a bug which would accidentally tweet the contents of your emails.
I think his concern is not about a bug tweeting private contents, but any kind of exploitable bug that might give access to his account in any way, form or duration.
There's probably no way around this, since any addition is a possible attack vector.
