Which is the prototype?
I got a funny request today, one that illustrates how cool our software development process is. The Lead QA on my project (let’s call him Blackbeard) asked me to put a watermark on my prototypes so he can tell at a glance that he’s looking at my prototype and not the actual working software. With multiple tabs open, it’s hard to tell which is which. And that is precisely the point! So I’m adding a watermark. 
We’ve also been asked recently to put real data into the prototype to aid in the user feedback loop. My fear is without the tell-tale fake data we put in, our users might run into the same problem!
Our process we’ve developed here at the Church has the Interaction Designer working out all aspects of the user interface and feature set in a high fidelity (XHTML/CSS) prototype before the developers ever see it. This way, stakeholders and users can see what they’re actually going to get, not some bloated bullet list of hard-to-understand functions and use cases.
We then sit and work closely with the developers and QA to make sure what we’ve promised visually is being fulfilled in the working software.
The downside, it seems, is when the actual app is too close to the prototype and everyone gets confused.