sports linguistics
Several recent Language Log posts have channeled a series of Tank MacNamara strips poking fun at the language of sports personalities. It's not clear whether the posters (Thomason and Liberman) subscribe to the tongue-in-cheek treatment in the comics, but my guess is that both would actually find broadcast discourse to be a worthwhile subject of analytical study rather than an object of ridicule.
I think there's some interesting observations to be made in this domain. On piloklok we occasionally track lexical-level phenomena in sports language, but I'd argue that the structure of the language of sports broadcasts differs in principled ways from other domains of language usage in every level of analysis - the discourse structure is pretty tightly constrained to begin with, but the syntactic, prosodic, and phonological structure could, I think with serious research, be shown to have properties adapted to the circumstances. Ferguson's work on sports announcer talk (Language in Society 12, 1983) suggests exactly this.
As for the apparent inanity that arises in on-air analysis (e.g., the stating of the obvious), it is essentially true that the parties involved are "paid to talk". Another way of framing it is that it's just in everyone's best interest to talk: this applies to the play-by-play and the colour analysts in broadcasts, whose goal is to entertain the viewer (try watching baseball with no sound to see what I mean), and to the coach and players in interviews and press conferences, which gives their organization good PR and the appearance of accountability. These reasons may provide enough motivation for the speakers to have a fairly high ratio of words to new bits of information. I don't know if there's an objective way to measure this, but I remain curious.