pks: I don't mean this in an abrasive way, but your analysis aroused my curiosity and I did some figuring. My calculations differ from yours, as do my empirical results.
I have a Puma-64. It has a 20" diameter impeller, rotating at 540rpm at full pto speed. This produces a tip speed of 2827 ft'/min, assuming my 22pto horsepower are sufficient to keep the impeller running under load with no appreciable slowdown. Why am I able to throw snow 50' or more despite not having the 4500 fpm that you posit?
Here, I think, is how. 2827 fpm tip speed translates to about 47' per second. The impeller is accelerating snow thrown from the end of the impeller to about that velocity (and snow not quite at the end to something less).
To attempt to be more precise, the velocity of the impeller arm 4" in from its end is 2262 fpm or 37' per second. Centrifugal force will tend to throw the snow out towards the tip of the impellers, but even disregarding this, the average velocity of snow coming off the final 4" of the impeller arm is still about 42' per second.
Assuming no wind, that my deflector is sending the snow up and out at about a 45 degree angle, and no loss of velocity due to air friction, the upward and outward vectors are each 29.7 feet, i.e., the snow will travel upward 29.7 feet while going out 29.7 feet during the first second of flight. (the upward and outward vectors are the legs of a right triangle with 45 degree angles, so their lengths are each the square root of 1/2 the square of the hypoteneuse of 42')
Since the acceleration force of gravity is 32' /sec/sec. (i.e. the downward velocity from gravity reaches 32'/sec at the end of the first second) the snow will continue to fly upward and outward for well more than a second before it starts to fall. (For most of the first second, gravity is reducing the upward flight of the snow by considerably less than 32'/sec.) It will then take well more than another second before it hits the ground, assuming that it takes about as long for the snow to return to the ground as it spent rising. Friction with the air will slow both the rise of the snow, its fall and its outward progress from the snowblower, assuming that these come close to cancelling eachother out, the snow should be airborn for at least 2 1/2 seconds during which (except for air friction slowing its progress away from the chute) it should travel some 59.4 feet or possibly more. ...all this with tip velocity of (only) 2827 fpm.
I realize, of course, that these are only approximations, because they do not take account of the actual effect of air friction, wind and perhaps other factors. I am no engineer or physicist. But the point is still clear; I don't think that you need to have anything like 4500 fpm tip velocity in order to throw snow MUCH farther than 20'. My Puma does not and it is certainly capable of throwing most snow between 50' and 100' depending on wind and snow conditions.
Needless to say, if there is an error in my reasoning, I would be most interested. BTW, it's lots of fun doing something wholly different from my usual thought processes. /forums/images/graemlins/grin.gif