I agree with the annoyance of everyone and their mother wanting their own app, but NPR seems like a big enough piece of a lot of folks entertainment/news lives that it makes sense. (I say this as someone who never listens to NPR.)
The question is whether the product is used frequently enough and can actually make use of a custom interface, compared to whether a more generic app would actually provide value. If I would use the generic app to listen to NPR 95% of the time, then it probably makes more sense to dedicate the app to NPR and just use a web browser for the other 5%.
Reversing XKCD, it's probably better to think of apps as slightly heavy and custom websites. There's really not much loss in having lots more apps so long as they're lighweight. Throw them all into a folder on your homescreen.
They have to play all sides. Replacing FM listening is like replacing oil: there's no one platform that big. Half of listening occurs in cars, and as the fleet turns over to TCP/IP listening over the next decade, NPR has to be there wherever. There will be branded apps, they will serve third-party apps and car specific apps, etc. Another audio provider (i.e. Pandora) could license NPR content and have access to the same APIs. I am not suggesting this is in the works, just a pure hypothetical argument. I haven't worked there since February but I did work on APIs and worked on this listening app a few years ago.
The question is whether the product is used frequently enough and can actually make use of a custom interface, compared to whether a more generic app would actually provide value. If I would use the generic app to listen to NPR 95% of the time, then it probably makes more sense to dedicate the app to NPR and just use a web browser for the other 5%.
Reversing XKCD, it's probably better to think of apps as slightly heavy and custom websites. There's really not much loss in having lots more apps so long as they're lighweight. Throw them all into a folder on your homescreen.
http://xkcd.com/1367/