That's an excellent point and one that I've hit on before. When a new technology comes into play, nobody really knows how to use it. The first initial steps are always some extension of something that we already know. The furor surrounding the development of a workable podvertising model is a great example of this. The initial steps were to bring web's CPM and print media's "per issue (downloads)" models into play. Now that the space is maturing a bit, other models are being derived that are focussed sepcifically on the distinctiveness of podcasting. (I get full points for the subtle Borg reference).
Check out Karl's roundup and see what you can add to it.