I couldn't work it out either so rather than storing a hyperlink, I just stored the file name. The path structure was stored separately. Then in the code I concatenated the path to the file name and used the followhyperlink method.
My application needed to manage 5,000 + endorsements for an insurance application and they were divided into to categories - admitted and surplus and within those two basic categories, they were divided by product. Your path structure may be more or less complicated but once you figure out the correct structure for storing your paths, the rest falls into place.