I know it's a different use case but the OP is talking about using a strong static message and having faith in the user to scroll below the fold...all the while as his website uses an animated popup box to fill the bottom right area of my screen...twice (on the iPad). So apparently, there's a case to be made for some kind of obtrusive moving elements...
As someone already said, I'm inclined to agree with the OP. But there needs to be hard data here. People often won't know what to do (I.e. scroll down below the fold) until you flash some hint in front of their eyes. Newspapers used to do usability studies on their print product quite often...and they always found that a huge number of people, no matter how compelling the story, did not turn the page. I can see a case being made that even if there is one big thing you want to get across to the user (such as a bug sale), there are often competing concurrent interests that would be missed below the fold. At least a carousel gives that item/module a fighting chance to be just noticed.
"People often won't know what to do" - do you have any data on that?
It is generally assumed in interaction design for web that the most instinctive action is scrolling. The concept of there even being a fold is also somewhat misguided, it like turning a page does not apply to a screen.
Sorry, "won't know what to do" is an inaccurate simplification of the problem. A better way to say it is that they may be ignorant of what's possible/happening.
Take a site like the iOS App Store (yes I know it's technically not a webpage, but same concept)...I go there to search for apps I want. I will almost never scroll down to the bottom because my first instinct is to hit up the search box at the top. In the time that it takes me to type a search in, Apple could show me up to two, maybe 3 featured projects that I otherwise did not know I would be interested in.
Could Apple have trained me to, once I hit the ios App Store front page, to scroll down to check out the list of featured apps and then conduct a search? Sure, but I'm at a loss as to how they could do that without encumbering the user (i.e. putting the search box at the bottom of th epage)
As someone already said, I'm inclined to agree with the OP. But there needs to be hard data here. People often won't know what to do (I.e. scroll down below the fold) until you flash some hint in front of their eyes. Newspapers used to do usability studies on their print product quite often...and they always found that a huge number of people, no matter how compelling the story, did not turn the page. I can see a case being made that even if there is one big thing you want to get across to the user (such as a bug sale), there are often competing concurrent interests that would be missed below the fold. At least a carousel gives that item/module a fighting chance to be just noticed.