* On Amazon, they did a quick-read of the top, and scrolled past a huge amount of info to get to the meat - the non-5-or-1 star review. Note that the one focused on is a 3-star, and the only non-5-or-1. Frequently, the 5 and 1 star reviews are fluff, so of course people skim past them to get to the more-perceived-as objective reviews.
* On JC Penny, were they just browsing, or looking to buy? I'd be willing to be that staring at sofas doesn't tickle many people's fancy unless they're seriously shopping. Sure, the fold applies to passive skimmers, but find someone determined to find something and I'd be willing to bet you'd see a significantly different eye-pattern. And note that the site isn't particularly heavy above-the-fold: JC Penny's site makes almost no money from people who aren't buying something, why should they worry about passive skimmers in their design? They're already past the front, attention-grabbing pages and into product lists, they're either buying or they're not.
* On the Zyflo CR site, note what content is at the top. Bulleted lists, images, bold, colorful text, etc. Eyecatching. And note how much they read further down the page, because it's all informative. The last bulleted list being skimmed is inconsequential once you realize what the list contains: things already said in the content above it. Other thin areas are excessively technical to most readers, so of course they weren't giving it as much information as the more generally applicable content.
The last site in particular strongly implies the opposite. Note how heavy above-the-fold it is, and how much was read further: http://www.zyflocr.com/zyflocr-healthcare.htm (yes, the article arguably "explains" this by stating that good AtF content encourages people to read further. It still contradicts the 80/20 theory, and seems to me that it implies FAR more that well-designed, easily-scanned information is more likely to be read.)