Christopher, I'm thinking exactly what your transcribers are doing, with the uhs and long pauses. Do they or can they notate when there is a change in tone within a sentence? I've noticed sometimes that a person will start a sentence, have a thought and change gears in the middle. It still looks like one sentence on paper, but is actually two distinct thoughts (and couldn't work as a sentence.)
As for shot descriptions, I do that myself when I log the tape, and I'll highlight things that standout, even with dialogue that might be transcribed later. The type of shot within an interview doesn't matter, only if there's unusable camera movement.
I've also done variations, with no transcription, but more detailed logging – where with an interview I'll write down what topics they are talking about, with timecode, and transcribe specific passages that come across well. In FCP I'll use markers and subclips to divide things up into specific bins.
One cool side benefit of doing translations – with the workflow I use – is that towards the fine cut stage, I end up with dedicated video tracks with subtitles using the FCP outline text generator. I can actually export those and come up with a word document with all the text and timecode. That way as I near a finer cut, I can basically export a "script". It's great to be able to read it and see how the cut I have is developing as a story or where there may be gaps.