The literature on this goes back to the original Whitney & Mehlhaff paper, which studied "the traumatic injuries sustained by any animal falling from a substantial height.”
Whitney, Wayne O., Cheryl J. Mehlhaff, High-rise Syndrome in Cats, JAVMA, Vol. 191, No.11, PP. 1309-1403, 12/1/87.
This study documents injuries and outcomes in 132 cats that had fallen two or more stories and were taken to the authors (New York City veterinarians) for treatment. Its main objective is to explain why and how most of these cats survived and to describe the nature of their injuries. The authors were surprised at one observation and speculated as to its reason:
... the rate of injury was approximately linear up to a distance fallen of approximately 7 stories. Surprisingly, these injury rates did not continue to increase with falls of >7 stories, and the fracture rate decreased. ... A possible explanation for this phenomenon ... is ... during free fall, cats have a unique ability to quickly minimize postural torque, rotation, and tumbling to maintain a feet-first landing position. An averaged-sized (4-kg), horizontally outstretched cat maximizes drag and achieves a terminal velocity of approximately 60 mph after falling approximately 5 stories. Cats falling from higher heights do not accelerate beyond this speed, but continue to fall at terminal velocity. It was surprising, however, that the fracture rate decreased in cats falling >7 floors. To explain this, we speculate that until a cat achieves terminal velocity it experiences acceleration and reflexively extends its limbs, making them more prone to injury. After terminal velocity has been reached, however, ... the cat might relax and orient its limbs more horizontally, much like a flying squirrel.
There exists a far better explanation and it requires no such (wholesale) speculation: the cats treated by veterinarians are not representative of cats who fall out of buildings. There is a high likelihood of a blatant selection bias: cats who fell but were uninjured were not admitted to this study; cats who fell, were injured, and survived had good chances of entering the study; and cats who wound up as bloody smears on the sidewalk would never have appeared. This alone casts doubt on all general conclusions about "high-rise syndrome" based on this study (and comparable subsequent ones).
Further evidence in support of this explanation is the authors' observation en passant that
The syndrome is seen predominantly in young cats...
These would have the greatest chances of survival among the general cat population. Older cats who fell generally didn't make it to the vets' office.
Given the obvious physical relationship between height of fall and shock of impact, we can therefore predict that the cats in this study who fell more than a few stories were the lucky ones, not the typical ones. Thus, there is no valid way to draw conclusions from trends seen in this study to any actual phenomenon.
Both this and a subsequent study report 90% survival. This, of course, is the survival rate among the cats who did not die before reaching the vet. (Kapatkin, Amy S., David T. Matthiesen, Feline High-Rise Syndrome, Compendium for Continuing Education, Vol. 13, No. 9, pp. 1389-1403, 9/91.) What should we conclude from this? Only that if you can get your injured cat alive to the vet within a couple of hours after its fall, it will have excellent chances of living due to the vet's skills in emergency diagnosis and treatment.
The only conclusion about high-rise syndrome that can be supported by this evidence is that cats who survive the veterinarian's treatment could have fallen from practically any height. The other conclusions that people have drawn--the "parachute" or "flying squirrel" theories, that cats will survive falls from any height, etc.--are not supported by this study (nor by any followup studies I had been able to locate in 2006 when I investigated this matter).
Afterward
Depending on your mood, you may find it either amusing or disheartening that pure speculation can quickly enter textbooks as received fact. A popular and well-respected physics textbook (Halliday, Resnick, and Walker, Fundamentals of Physics, 7th Edition, Part I, p. 123. J. Wiley & Sons, 2005) not only incorporates the Whitney & Mehlhaff explanation wholesale, it amplifies it and makes it more plausible by adding details that nobody could possibly have observed:
According to calculations, ... a cat must fall about six floors to reach terminal speed. Until it does so, [the gravitational force exceeds the drag force] and the cat accelerates downward because of the net downward force. ... Because the cat also senses the acceleration, it is frightened and keeps its feet underneath its body, its head tucked in, and its spine bent upward, making [the effective cross-sectional area of its body] small, [the terminal velocity] large, and injury likely.
However, if the cat does reach [the terminal velocity] during a longer fall, the acceleration vanishes and the cat relaxes somewhat, stretching its legs and neck horizontally outward and straightening its spine (it then resembles a flying squirrel). These actions increase [the effective cross-sectional area of its body] and thus also, by [the drag coefficient equation], the drag. The cat begins to slow because now [the drag force exceeds the gravitational acceleration] (the net force is upward), until a new, smaller [terminal velocity] is reached. The decrease in [the terminal velocity] reduces the possibility of serious injury on landing. Just before the end of the fall, when it sees it is nearing the ground, the cat pulls its legs back beneath its body to prepare for the landing.
It's a good story, but it's merely a statement of faith, not science.