The reasoning behind linking to 148apps and the like was because most app buyers read reviews before making a purchasing decision - so this was meant to help in the use-case, as well as standout above the iTunes search which doesn't have independent reviews.
But, I see where you're coming from and will definitely think about shifting to that track.
Maybe mimic Amazon's review system? Get all the reviews from iTunes and augment them by letting people say helpful/not helpful, and then showing the most helpful highest rated review and the most helpful lowest rated review, etc.
Basically create a landing page for each app and then optimize the )(*!@# out of it.
The reason I think this is the right approach in your case (vs. Google) is because your universe is very contained. There's a single, canonical database of applications you need to index (for iPhone apps). Google has to index the whole web, so creating verticalized sub-products is hard to do without knowing what's important and what isn't.
Google is obviously moving in that direction, with maps, reviews, music, video, etc.
This might be different for Android vs. iPhone. I'm not very familiar with the Android ecosystem.