Looking at just winners isn't good enough to determine if a course is a "good test or not".
That would be like saying Jonathan Byrd is the best player in the world because by percentage, he gets up and down from the sand better than anyone else. Of course maybe Jonathan Byrd plays easier courses and therefore has better opportunities to get up and that may bias his sand-save percentage.
Statistics are great, but they are normally biased. Even the final results of a leaderboard don't always show accurately who was the best player that week. They just show who scored the best. (Now to us golfers, we understand that a score is the bottom line. You determine how well you play by your score.) But of course "we all know that there was that time when we were playing great golf, striking the ball pure, chipping it stiff, and putting it great, but a bad bounce on 18 puts us into the water. We salvage triple bogey after a few, half-hearted shots, and there we are, a score 2 strokes worse than our personal bests." But for 17 holes you were playing as good as you have ever played.
So I think we should look at a final score or the final standings of a tournament as more of a statistic or measurement of how you played or who played the best, not the ONLY determinant of how you played or who played the best.
Oakmont is obviously an amazing golf course and an amazingly hard golf course. I think it is worth noting that 2 of the top 3 players in the world finished in a tie for 2nd.
I think there is something tangible (proven by statistics) and intangible about who are the greatest players in the world. They just get it done. I think majors do a greater job than any other tournament of allowing the greatest players in the world separate themselves from mediocrity.
If you cut a cup in the center of a perfectly flat, non-hazarded circular green, on a par 4 that is a straight away 380 yard, 50 yard-wide fairway with no bunkers or hazards and short rough, more than likely, you'll see a player or two make eagle, A BUNCH make birdie, a fair amount of pars, and a few bogeys.
Tiger will make birdie, but so will the 90th best player in the world. This hole doesn't do a great job in separating who is more outstanding.
To answer Mr. Young's original question, we should not tailor our tournaments to achieve a certain winner, or set of winners. You should tailor your tournament to require outstanding performance if one wishes to score well. Of course in doing that, the better players will naturally rise to the top, but it isn't about doing just that. Ben Curtis wasn't playing like crap the whole week in 2003 at St. Georges. Tiger was in the hunt down the stretch. More good players were in the hunt down the stretch. Colin Montgomery finished 2nd to Tiger in 2005 at the Old Course. It wasn't like the world #1 was the only good player at the top.
Having a well-known golfer is a result of him being close to the lead down the stretch often. This produces a high ranking the the World Golf Rankings and more notoriety. The more the player is close to the lead, the more relaxed he is in this setting and of course, if he is close to the lead, his is obviously playing good golf as well.
It is kind of a circular argument, maybe reciprocity might not be a bad word for it, but it also goes somewhere, maybe it is more of a spiral type argument, and hopefully I approached the point.