The threads premise makes perfect sense. To build a great course on a bad site may be the ultimate test for an architect.
Yes.
But to sum up the way of looking at things, you have to define a great site / bad site in to two sub-sets:
1. Size / shape of land
2. Soil and natural features
For number 1, a great site can yield a very difficult routing if it is wild or small or strangely shaped or have a load of environmental restrictions. A bad site can yield a very easy routing if it is big with next to no natural features: You just blow it up and route what you want.
For number 2, good soil and good natural features (given a good and easy land parcel) is always easier to make a great course than if given poor soil with no features.
After this, you then have to define what makes good architecture. Again let's divide that in to two:
1. The detail / aesthetics / strategy
2. The technical side of architecture (drainage, irrigation, engineering, science, cost & contracts)
An architect on a great site can sometimes make a great course by excelling at point 1 but not necessarily point 2.
An architect on a bad site can only make a great course by excelling at both points 1 and 2.