So-called "simultaneous" translation is often done that way -- the translator has a written copy of the speech to work from, but also listens to the speaker live in case there are last-minute changes.

It makes no sense to have someone else read the transcript out loud. Simultaneous translation is challenging enough -- I've done it before and found it one of the most nerve-wracking experiences of my entire life -- without having two different simultaneous voices to handle.

This explanation doesn't make sense.