I believe Pat's answer lies in the laws of scalability. I'll explain...
As a given shape, say a triangle is stretched from two points (a & c) the relative angle to the third point (b) decreases. . . .
Thereby minimizing the initial effect / relationship of the strategic features.
Jim, I finally got to this thread (I had temporarily given up on Patrick's and Rich's approach to this issue.) Interesting explanation of scalability, length, and its effect on strategy. I had been contemplating a thread on a similar topic but would not be capable of explaining it as well as you.
As I started reading I thought I was inclined to agree with your analysis, but your approach and schematic actually raised some questions in my mind. I always try to think of strategy from the green back to the tee, and if I do so in relation to your diagram, I find the crucial angle unchanged.
As you accurately depict in your schematic, courses are usually stretched from one point (the tee) with the other points (the green and features) stationary. While this closes the angle of the tee shot, it doesn't alter the angle of the approach. If the proper angle of approach is crucial to success on the hole, lengthening the drive would not change this. The golfer would still have to get it to the same spot, albeit from further away. Let me try to draw it.
_________________________________________________o.b.
a
T(2) T c xxxx G
x
b x
Let's pretend this is a long par 4 with lots of width. T(1) is the original tee, G is green, xx's are features. Let's also pretend that because of severe green contours and/or the front feature, the ideal angle of approach is from Point a, up against the o.b. Point b is the second best, but it is also protected by a hazard. Point c is the safest drive, but least advantageous approach.
When the tee is moved back to T(2), The driving angle decreases, but the ideal angle of approach stays constant. If strategy is dictated by the angle to the green, then the strategy would appear to be intact.
That being said, I do think that added length might sometimes curtail strategy, but for slightly different reasons. As the tee is moved back, the margin of error (or the degree of error) for the tee shot decreases, assuming it will be hit further (the reason for lengthening.) This brings the o.b. and other hazard more into play. As a result, the strategic golfer has a different risk/reward calculation from the tee, and there is a greater chance he will err on the side of caution, and play for c. Taken to an extreme, a and b might be so risky that they are no longer viable avenues at all.
Just thoughts. Feel free to tell me where my logic falters.