Fun is a great point, and the few who advocate for some puritanical life test of values and morals seem to miss the point.
Why do most of us play golf? To get away from the stress of every day life. Why would we want some life or death challenge? As I have said before, 90% of the time, 90% of golfers want a suitable challenge (who wants to play putt putt and brag about their score of 45, for instance" think that challenge is suitable when they shoot an explainable score near their average (i.e., well, I put three in the water on 12, so I didn't break 90."
And, yes, some potentially "great" architecture does get eliminated when a gca tries to average out all those feelings in trying to be suitable challenge to more golfers most of the time.....but I think we could also argue that some great (but punishing) feature really isn't all that great if it hurts most golfers. I think Mac said something similar. To Garland's point, traditionally, a good design is somewhat flexible, in that rough can grow, greens speeds can increase, and pin positions can be moved, to increase difficulty on specific days to some degrees.
If playing for more than you can afford, is it you sweating, or your wallet? Even so, all but really bland designs have a way of separating similar golfers by a stroke or hole, without separating bad golfers by dozens of strokes, at least in most cases.
Short version, isn't the greatest good for the greatest number of golfers (and perhaps others, i.e., environmental issues) a good criteria for assessing golf architecture?