George,
I asked a similar question in a thread not too long ago. And received some interesting replies. As I recall the overriding message was, course difficulty and architectural merit need not be coupled. Having said that, I think there is a scatter plot out there somewhere which would show a correlation between course difficulty and perceived quality. I think there would a level of difficulty where the "quality" y axis would flatten out but early on in the "best fit" line for the scatter plot, the relationship would be linear.
I realize that slope is an imperfect measure as it really is just a subjective value by which course difficulty for bogey players relative to scratch players is quantified. And although it's done by raters using a set of predetermined criteria, the formula uses those criteria to predict what the score would have been. But instead of having a 100 bogey golfers go play the course, they estimate. Seems at least partially reasonable. We use use formulas in everyday life to predict events (weather, retirement funds, the NFL draft! etc). So why not predicting a bogey golfer's round?
I am doubtful that a course would be considered architecturally great if it had a low slope. If you want to go to the top of the architectural mountain, here is a top 10 listing. The slopes from member tee equivalents (around 6500 to 6700) are as follows:
PV 153
ANGC only Old Tom knows for sure
CPC 140
Shinny 138
Oakmont 134
Merion East 149
Pebble 136
Winged Foot West 134
Sand Hills not rated near as I can tell
Fishers Island 143
I'm not advocating slope as the best or even a good measure of gca. But I am saying a smart statistician somewhere at MIT would be able to demonstrate they're correlative. Put another way, an architecturally respected course will probably be considered hard enough by bogey golfers relative to scratch players to have a higher slope than a course that is not regarded. The converse is not necessarily true - a course with a high slope could be a gca train wreck.