If a directory, especially one with a local component, is or will soon be expected to play and compete on the same field as search + G Place Pages I'm hesitant to believe that the future is one where miminalist directories will remain viable, in any sense of the word.
This one's chicken and egg all over the place. I think your instincts about depth of data and differentiation are correct.
The question is whether it's possible to start a new directory that is likely to be helpful enough to a business to motivate them...
(a) to list at all, and...
(b) to fill out your form.
Google Place Pages are inescapably there, and those who don't claim and maintain them do so at their peril... but most businesses don't have a clue. While Best of the Web has been giving their own local place-type pages a try, I also see a great many blank categories... which is too bad. Maybe they need to seed the listings more to get things started.
But there are also a huge number of listings on a great many late-to-the-game local restaurant guides, professional guides, etc, which have been seeded and are nothing but stubs. How many "be first to add your review" listings have you seen on review sites? How many invites to the latest professional or personal networking site have you deleted? In truth, the abundance of these sites has unfortunately turned into a major PITA.
So, you've got that hurdle... attracting and motivating those who either have incomplete listings or no listings at all.
If you can do that... guided form fields, I think, are a good way to go. I've suggested them in lieu of (or in addition to) a blank description window for several clients, in a variety of areas where User Generated Content is being used or considered, and present UGC is lacking or deficient. All clients I've suggested this to have hesitated because they feel that large form inputs will discourage response, or because it's too much trouble to set up such a system.
I recently responded to an email customer feedback request on an ecommerce site because I really liked the product and because I wanted to evaluate the experience. The onsite input page presented a sequence of short input fields that had if/then type inputs... somewhat interactive... so I was initially presented with a smaller number of blank fields then I ended up with. The task was short enough to be manageable in this fragmented world, and the resulting description, which was shown to me before I submitted, was surprisingly OK. If I liked or disliked a product enough, I'd do it again.
On a directory site, you might want to give the user the option of doing things in stages, much as Linked-In does with its profiles. I know I will never "finish" that profile, but I might get around to adding to it.