You missed the point as far as gene therapy goes. They didn't use an modified adenovirus to piggyback the gene that they wanted to add. This technique could work for other cells along the G.I. tract from your mouth to your behind.
If you were a mouse. The diabetes researchers who I've heard speaking on the topic are quite clear in their comments about the difficulty of transferring mouse results to humans. It'd be great if this transferred, but I'm betting this is another example where mouse results are good for mice, and nobody else.