I think the popularity will come down to the same three factors as usual: variety of content, price, and ease of use. The direct connection to Amazon, while limiting, could be good for ease of use, but I don't think Mobipocket has enough content yet (and right now, you don't see Mobi ebooks from the regular Amazon site -- and the Mobi site isn't so easy to use). That leaves price -- Mobi prices aren't any worse than the rest of the market (or not much worse, anyway), but not really competitive with paper books, either. The price would have to come WAY down for anyone to be willing to pay US$400-500 up front for a single-purpose device. As in, practically free. But since EVDO will have its own cost, I don't think that's going to play out.
The only way I see this working is if they get a couple of major textbook publishers on board-- at greatly reduced costs. I just paid nearly US$70 for ONE textbook. If textbooks were something like US$10-20 on the Kindle, a full-time student in the US could recover their investment in the first year. But the only way I see the textbook publishers going in that direction is if the DRM is time-limited.
