The article title demonstrates one fallacy. The people read "coffee and tea" and assumed this has to do with caffeine. Buried in the slashdot thread some people are celebrating that the coke they drink every day ought to help them.
Somehow I don't think this is about caffeine, but about anti-oxidants.
The study is said to have been conducted by the National Institute of Diabetes and Digestive and Kidney Diseases and Social & Scientific Systems. Unfortunately I can't find the study on either website.
The study was published by the American Gastroenterological Association journal Gastroenterology. I see that in the November 2005 this is listed as a forthcoming article. And, the article is available here but requires registration with the web site.