Officially, a URL-only entry means that G knows the page exists because it's found links to it on other pages, but it hasn't crawled the page itself yet. But, there have been problems reported by webmasters with regard to having known pages go URL-only, with associated ramifications. I just have the regular 5-10% of pages sometimes show up that way, varying, and for reasons unknown, so I can't give you the low-down on the drastically affected sites. Doing a site:www.webmasterworld.com in G should find you a lot more info.
Are these pages that don't show any description database driven?
A couple of months ago we had the same problem with 90% of our websit, due to the fact that our dynamic pages use Isapi rewrite. The implementation was of such a poor and awful quality that the construction of the pages took so long that Google always received a timeout when spidering our pages. The time to parse the pages might be a factor.
You might sign-up to the Google sitemap project and use its statistics to analyze errors returned from Google when spidering your site.