I've asked this question before and never really received a satisfactory answer:
How does one really truly and accurately evaluate a "great" site? It seems to me that often people base the site on how the final product turns out, especially with regard to the architect. If he's a "minimalist" and the course turns out great, obviously he was working with a superior site (others' opinion, not mine). If he's someone who tends to move more dirt and a course turns out, shall we say, less than special, well, obviously he was constricted by the limitations of the site.
As a corollary, how dependent on the site is an architect when producing a superior course? I know setting can make things extra special (think Cypress Point or Sand Hills, or any of the many oceanside courses), but I continually come back to the example of Oakmont. This land is, at least on a superficial level, no different from any other land in western PA, yet there is only one Oakmont. If someone can explain to me the difference in Oakmont's property versus the rest of western Pa, I'd love to hear it. You could probably say similar things about Lehigh. Its hilly, somewhat severe land would be right at home in western PA, yet it is a much better course than most we have over here.
Look at something like Doonbeg - dunes galore, but since the course has not quite lived up to expectations (at least, that's how it seems to me), now people are saying that the dunes were too severe. Is the property really that different from Ballybunion's? I ask this as a serious question, I've never had the pleasure of visiting the Emerald Isle.
I think site selection is far more difficult and uncertain than most think, and I think it is much more incumbent on the architect to truly maximize a property than many believe.
If great courses like Oakmont and Riviera can be built on land like the land they sit on, I'd say there are a helluva lot of courses out there that sit on sites that were not taken full advantage of.