I, too, am restlessly awaiting audio. What has happened to the Swahili team? They don't seem to have been active for months.
If all the audio is done and labelled, and merely needs to be matched with the corresponding sentences, could one not volunteer to help them out?
Other incubator updates have strongly suggested to me that the assigning of recorded audio to sentences is a matter for contributors. It's certainly an admin job with TTS audio, but Swahili does not have this.