They built it, they own the ecosystem, so they get to decide.
Don't like it? Write a webapp instead.
Any time you implement a feature for a product you have to do it in a way that can't possibly please everyone. If you let your developers implement alternative versions, it's conceivable you could get close to pleasing everyone, which makes your product better because it then covers obscure use cases that you can't afford to deal with. It becomes as if third parties are developing the product for you for free.
Hypothetical example: What if apple is already working on a better podcasting app that works on the iphone, yet they _don't_ reject any apps?
Then, when their podcasting app comes out and kills off the competitors (usually because it is free--but also because they write damn good apps), then some author is going to bitch about how apple targeted them and killed their app. Sounding like that whiny sandvox guy doesn't do anyone any good.
Maybe... just maybe... this is the better alternative.
Personally, I'm going to write the quickest/smallest app I can that gets the 1.0 beta 1 idea into form. I'll apply with that. If I get accepted, great. I'll pump out a lot of releases getting it to 1.0 final. If not? I didn't lose that much time/effort.
It was never 100% clear that Apple would bar that. It still isn't.
personally, I don't see whats particularly troublesome about the current policy. Granted, I'm focusing moreso on cooking up apps that are genuinely distinct in functionality to minimize the need to focus on marketing to begin with.