To me, it's kind of like Yelp - sometimes I want to just know what has the best food, and don't care about the service. Sometimes I want to know both.
What I'm still trying to figure out is how to implement this from a user perspective - it's difficult enough to get a user to provide a review, let alone multiple reviews for different purposes.
Perhaps the genii of YC have some good ideas...
People like seeing information where its supposed to be, and ads where they are supposed to be. Part of what makes Web 2.0 websites so great is that the ads aren't in-your-face annoying. Here, your ads are messing up the user experience in an attempt to get in front of the user.
Some people will want to do a quick review, others will want to leave a comment, and still others will want to fully explain how it helped them or what they liked/disliked. This way you take into account various user participation levels with the app the user is reviewing. The final rating could be a weighted average of the three types of reviews, with in-depth and commented reviews being rated higher. Knowing your review will count more and be more important if you put more effort into it is another psychological tool to get people to write about the apps they care about.
I thinking building on this would really up the usefulness of your site immensely, because people want to know what's out there when they're looking for tools. Rating is secondary to someone looking for tools. Instead, they need to know what tools exist. Building out the categories more is a suggestion, as would be providing your own content (editorial) on how well apps in the same category work compared to each other. For example, if you're looking for a commenting system for your blog, what are the pros and cons of the major offerings. I understand this might run into a problem of what is a web app and what isn't, but people love to have someone else do the work of finding them a solution. That involves more than a one line description and a rating.
Rating is good to eliminate the lowest bottom feeders, but it doesn't help you choose between two five-star programs. Original editorial content helps people by providing more information, as will user comments of course once you get a mass of users.