PDF is a guarenteed-to-render-identically-everwhere format. Its input is the postscript file that is sent to the printer. No need to have client side fonts, office-suites, directory hierarchies, etc. Just pick up Adobe Acrobat Reader and you'll see it the exact same way everyone else does, the exact same way you printed it, just 1/100 the size.

Every browser has a different process to create the image you see on the screen from web pages - there's a lot of development taking place now in modifying even that html data stream as it comes in, with things like Greasemonkey, Noscript, Platypus, Adblock. The goggles that we see the internet through are definitely not perfectly clear, and often involve annoying hacks to try and get pages looking the same on all browsers.

By the way of feature creep, Adobe Acrobat has lost its way, and now sluggishly loads completely unnecessary features like audio and shockwave and kitchen sink that it can take minutes to open. Downloading the free Foxit PDF client is the immediate fix for 95% of files(hundred page docs browse faster than notepad), with the long-term solution being adaptation of some type of open XML-based standard like ODF (correct me if I'm misapplying this), which should take at least a decade. The public sector doesn't let go of bad formats quickly - CNN and BBC are STILL using realplayer, after most of us uninstalled it in the 90's for good.

I don't really think we NEED stories in PDF, but the webpage table-based formatting is getting challenging to read at higher resolutions (telling the browser to read text bigger and zoom in on pics messes up formatting on any browser other than Opera). I hack away at the screenside cruft with Platypus when I have a problem, personally.

I use PDF mainly to print out longer articles I want to study (images onscreen make that difficult).

On Linux/Unix xpdf is a fast and mostly capable alternative to "feature loaded apps". They may have versions for other platforms. Evince on Ubuntu Linux is even better.

ciao,
Bruce