I've always thought that any given shot is best if the shot pattern is not strictly dictated, but at least strongly suggested. As I have mentioned before, in working with tour players, only a select few great ones really worked all the shots. JN and others generally prefer to perfect one shot type and practice it to perfection. Faldo was one who went with what the architecture dictated as the highest percentage shot for success/lowest chance of failure.
Those type of players assess the wind (blows left, play right to left), lie, left lie suggests right to left, and situation of target, i.e., green angles left usually presents the most favorable target on a right to left shot pattern. As Jim Colbert once exclaimed, "If the wind is left, the target is left and the lie is left, smarter guys than me, but I'm pretty sure I am going to hit a draw."
As opposed to Kyle's post above, shot makers prefer all (or most) of the signals suggest a shot, then they hit it. Colbert likened it to coming to an intersection that had both a stop sign and green light. Which do you think you should do? Of course, some architects think its a good idea to present a mix of signals to confuse the golfer, add doubt, etc. And yes, good players like the signals to align because it does clear their mind and allow focus on execution.
Aside from left to right and right to left shot patterns, other players have stressed how they consider high and low (pretty simple, really, tail wind = high, headwind, sidewind = low) ground contours (look for the speed slots for more roll, or hit into a counter slope if too steep cross ways) and high spin/low spin, sometimes trying to reduce spin into the wind, because it tends to add it, but then sometimes reducing it downwind to "chase a shot" all the way to a back pin.
Of course, around the greens, a variety of sand, rough, mounds and fw grass areas lets the short game shot maker with different strengths shine a few times per round.
I consider a shot makers course on where nearly every type of shot is strongly suggested on one hole or another, and through the course of the round each is the preferred shot with the highest chance of success. Thus, over 18 holes, the golfer with "all the shots" should have an advantage over one who doesn't, at least over the course of a competitive tournament or season. It doesn't have to be the arbiter of shot making every single round.
Lastly, the details count. The question becomes how strongly to favor a certain shot vs. what to do about the vast number of players who don't possess "all the shots." A typical example is a hole with a green 60 feet wide and 100 feet long, angling 30-45 degrees left, suggesting approaching from the right side of the fw, and usually with a sand bunker/hazard left side. A draw will best fit the green shape (the long dimension angling right allows more room for error short and long, vs. coming across the width from the right side, in this case.)
I would typically add containment back left as an extra inducement to come in from the right. But what about those coming from the left? Do you screw 'em, really creating the advantage from the right, or arrange things to help out the mere mortals coming from the other side? I can tell you the potential for "golfer on architect unpleasantness" often stems from situations where a golfer feels they have no shot.
Typically, a green angling left is higher at the back right, to promote visibility, so that golfer is probably at least has a 2% upslope to help stop his shot. That seems fair enough, giving them a way to hold the green on a short depth target. That same upslope helps the golfer from the right side aim far right and use green slope to roll a shot down to the hole, allowing further room away from the sand bunker left. If the wind is right to left as per above, coming from the left it may straighten the fade and maybe even add some spin, increasing the odds.
My debate is always what to do with the greens surrounds on the right side. Choices (from least to most forgiving) are additional sand or other hazard, a fall away slope that kicks long shots away from the green, a level open area leaving a reasonable chip for up and down, or containment mounding similar to the back left. Of course, there is no one blanket answer, even conceptually. My tendency to add hazard would be more likely on a shorter hole, trending up to more containment on long iron approach shots.
Conceptually, i
f the shot is 170 yards, they probably need 17+ yards depth, and have about 20 yards to stop the ball, a max of about 15% room for error, not as much room for error as from the right, where there is 33 yards of depth to accommodate the slight miss, which is 50% margin for length error. Adding hazard or a fall away rough bank makes the shot from the left of the fw, presumably with a fade about half as makeable (see math above) translating to the "half stroke penalty", perhaps? Maybe more.
Adding containment makes a miss long from either side a chip from just off the green, probably negating the problem of missing long, while maintaining the harder shot from the left, because you can't come up short. So what, a quarter shot difference?
To some, narrowing the difference between a shot from the proper side and the other side to something negligible is good design. It lets you play the hole any way you want, with only minor inconveniences. To others, a shot makers course would really differentiate the difference between preferred shot and others with greater penalty. And, in a few extreme views, if the penalty for hitting the wrong side of the fw is too severe, it is really penal design masquerading as strategic design, because there is really only a small target to hit off the tee no matter how wide it may look.
Obviously, getting that balance right is pretty important in creating a shot makers course, at least IMHO.