Well, if you say so :-) ... it's just that you can do all the interactive stuff with Javascript/CSS on the web, too, and not only mobile dev users would beneft. Maybe the op finds a way to also publish this information on the web somehow and esp. in the ml help system.
Personally I think an app like this would be useful
if it was well done. Here are some of the benefits I see:
1) Works offline (unlike HTML/web). 95% of my photography is done in foreign countries or remote areas where I don't have Internet access.
2) Could be far more in-depth than the on-camera ML help. Eg tutorials, example images/videos, hyperlinks, animations.
3) Possibility of an improved UI due to touchscreen & sensors.
4) As others have mentioned, useful for studying without the camera present or while the camera's on a tripod.
5) Preserving camera battery life (important to me when in remote locations without the ability to recharge).
6) Future possibility to allow some camera integration ala DSLR Controller.
I'd think of the ML on-camera help vs this app as being analogous to the Canon camera manual compared to a book like
http://www.amazon.co.uk/Canon-DSLR-Ultimate-Photographers-Workflow/dp/0240520408 (note that I have no idea if that book's any good, I just chose it randomly as an example).
Having said all that, I strongly agree with the general sentiment that it is far more important the in-camera help is brought up to speed as a priority, or if this app is developed it is important to be able to somehow keep the app's content in sync with the in-camera help and with the content on the ML website. At a minimum I'd hope that the ML website's userguide and the app's content was based on the same source content - the app just provides an offline version for those who might require it.