Download URLS

I'm wondering why disknode uses its own path (?q=disknode/get/6/document.pdf&download) as a redirect rather than simply the exact url.