It's not primarily about speed or even the mentality of the purchaser: It's about giving the designer of the software (and the books) the power to create a better user experience. If the books are really as good as Gruber claims, then that will lead to sales.

As long as there are non-trivial differences between hardware platforms, there will be room for native apps.

I honestly don’t see why native apps would require more effort and commitment than web apps, at least on devices with app stores of some kind.

Native apps can use some app store infrastructure where you enter your details once, websites would have to rely on PayPal or other services and you would have to enter your details over and over again. Paying in some app store has less friction. I can kind of see your point for free apps, searching some app store for an app and downloading it kind of seems more complicated than googling it and taping the link, I can’t (yet) see it for paid apps.

I'm more frustrated that no iOS e-reader solution accommodates extremely large libraries, allowing you to easily sort/filter by author, genre, series, etc. (Stanza comes closest.) If we have gigs of storage, why not carry around every (e-)book you own?

This situation is doubly bad with books as native apps, even with "folders", which still have caps and how many entries they can hold.

I don't see any reason why PPP's application would appear (on the outside) any different than any other eBook reader app.

I mean, I guess from the sounds of things it's up to them - they could deliver their books as apps, making them first-class citizens on your home screen, but that would get annoying fast.