Edward Tufte posted an essay and video with commentary about the iPhone. Tufte is complimentary, for the most part, about the interface being predominantly free of administrative debris, but makes suggestions for how the iPhone can make better use of its screen resolution. As expected, his argument calls for more data density when possible.
He notes that the iPhone has elegantly removed administrative clutter in many places, using the information as the interface. In reference to the iPhone's design of the weather widget, he says the following.
To clarify, add detail. Clutter and overload are not an attribute of information, they are failures of design. If the information is in chaos, don't start throwing out information. Instead fix the design. And that is exactly what the iPhone has done.
He makes the argument that the "chartoony" appearance of the stock market widget does this very thing—removes information that could be useful, given the resolution of the screen and the zoomable UI. His position on showing more data prompted some debate about the efficiency of more versus less in this specific context. There are many great responses by commenters that justify the clarity of the large type and the choice of only essential information in the iPhone, including the great retort by Chris Fahey.
You are neglecting the fact that iPhones are *mobile phones*, designed to be used primarily by people on the go, or by people who are otherwise occupied. The cartoony UI screens are designed to be usable by people who are walking, talking, riding on a train or bus, waiting in line, bored in meetings, and (unfortunately!) while driving.
Typical iPhone usage lends itself well to the information-thin designs you criticize precisely because it does not attempt to do more than deliver the most important information in a heartbeat. The “image resolution” style of information design you advocate is great for someone using an iPhone while sitting in a comfy chair with lots of time on their hands, or for someone who posesses no other information platform (i.e., no desktop or laptop computer). But for most users, they will use the iPhone to informally keep their finger on the pulse, and use their main computer to actually think about and analyze data.
Fahey's points touch exactly on the thing that Tufte is missing in his solution—discussion of iPhone users and the context in which they use these interfaces. You might find more data on a PDA or any other smartphone out there, but if you test those other data dense interfaces in real world scenarios against the iPhone, I would bet that the iPhone comes out on top in terms of efficiency and usability.