Zig, while I absolutely agree that there are loads of sensationalist and rediculous claims floating around regarding the 4S image quality, I think this test itself is actually a pretty valid one. But, it's valid only for a very select aspect of photography. The problem is related to the conclusions other people are drawing from this test. Because this test shows the 4S stacking up well against the 5D2 in very specific circumstances, they conclude it will perform just as well under all circumstances.
As you point out, comparing your Rav 4 to a Ferrari at a set speed of 80 mph is completely ignoring some of the main reasons for buying a Ferrari. But I would argue it's still a valid test of two cars driving at 80 mph. I would actually be interested to know what a Ferrari is like to live with when being driven the same way as a regular car. You just need to be careful about the conclusions you draw from such comparisons and remember that outside the confines of this test, the Ferrari offers a whole new world of opportunity that the Rav 4 doesn't.
The original video test of the cameras ignores such things as low-light performance, the ability to achieve shallow depth of field, provide full control over shooting parameters, and all those other things where the 5D2 would mop the floor with the 4S. But I don't think the testers were trying to trick the viewer into believing a 4S would replace a 5D2, I think they were simply examining one particular thing in detail. They were actually interested in how a Ferrari feels at only 80mph.
But aside from the "5D2 versus 4S" competition, what the video does show is that the 4S can handle certain lighting situations as well as you'd hope for, it does a decent job of reproducing colours, captures a reasonable dynamic range, estimates white balance pretty well (it's a bit warm in this example, but in my experience the 4S actually seems to do a great job with white balance), has good enough optics to justify shooting HD video, and does a remarkable job when you think back to the video quality of phones from just a few years ago.
These are all things which we take for granted in modern DSLRs, but they are still issues that haunt most camera phones today. In this respect, I find the video interesting, just as I find the iPhone 4S a truly useful camera. I would also have loved to see a third test camera included in the comparison; a previous-generation camera phone.
Here is another iPhone 4S versus DSLR comparison which I think is a bit more comprehensive and well rounded, although it's more concerned with still images than video.
And
this is yet another comparison of all the iPhone models compared to a Canon S95 and 5D2 for reference. While the OOF photos from the iPhone 3 and 3S aren't really fair for the purpose of comparison, the other photos are interesting, particularly from the S95.
I am a bit annoyed that most people seem to constantly want to compare the 4S to a DSLR, which they clearly aren't going to replace any time soon. A much better comparison would be to some of the more popular small compact cameras. I would love to see more comparisons with the Canon S95 for example. That's the market that camera phones are going to be eating into, and that's where some real meaningful comparisons can be made because both categories offer similar features and suffer similar limitations.
But regardless of any specific details or test results, there is one important message coming through by the presence of all these comparisons and discussions.
The time has come where phones are being seen as legitimate photography tools.