Peter,
I agree with Lou Duran about the long holes with/short holes against, as you do end up with a lot of driver/6 iron holes. Sorry Mr. Colt! Actually, we try to box the compass with not only par 3 and 5 holes, but also try to balance long and short 4's as well, so the golfer is likely to have at least one really long par 4 in any wind.
Our feature design respects the prevailing wind in the main golf season as well. It's easy in the midwest, where the summer is southwest winds, and fall/winter is northwest. A north facing hole can almost count on having a left to right cross wind of some type. In Texas, it's a bit harder, because summer winds come from the south, winter winds from the north. Our only saving grace is that it is usually too cold for a Texan to play this time of year when the wind comes from the north.
We design features to account for the wind, usually canting the green or fairway landing areas with the wind - ie a left to right wind, we - as often as we can - angle the green slightly right to better accept the likely shot (a good player will use the wind, and the ball will simply drift for the average player). Of course, we always follow the natural ground features, and if the wind will usually blow right, but the ground angles left, we angle the green left, but flatten out the angle and simplify the green shape to present one "blob" of a target, figuring the golfers will use different shot patterns.
We also design greens by giving more depth and less angle to the line of play to a downwind green, reasoning the wind reduces backspin, while we might elevate the green a bit more, figuring few will "run it up" downwind.
There is a short 4 here in Fort Worth that plays downwind, with a creek in front and a shallow Bermuda grass green. It's amost impossible to hold the green most days. Had I designed that hole, the green would have fronted the creek, but been deeper, allowing a shot to the back of the green. When the pin is in front by the creek, players would still have the challenge of getting close, or avoiding a downhill putt, but could hit some part of the green.
Conversely, into the wind it would usually hold as designed, and we would have probably done the green as the architect did do it. When a green is usually into the wind, a shallow green and frontal hazards (for moderate and short holes) is better, because the wind helps "stick" a shot. On longer holes, where there are no frontal hazards, we usually lower the green, allowing more runup against the wind.
Of course, we widen everything if we feel there will be a consistent crosswind, unless we want a really, really, tough tee or approach shot.
As for hazards, we use the wind to create the most temptation, if possible. Generally on tee shots, I place the the hazards on the up wind side of the landing area if they encroach into the fairway.
In a downwind situation, I place hazards in front or the landing area, to create carry hazards when the golfer has most confidence he will get the distance (as you mentioned in the REdan hole), pinching bunkers beyond the landing area in head winds . Directly into the wind, a hazard at 300 or so yards creates the "I want to get as close as I can to have the shortest second' dilemma", tempting the player to play aggressively close to the hazard, when downwind, that type of hole would be a no brainer to stay well back.
For flanking hazards in the landing area, I usually put on the downwind side, so the player can aim somewhere out over the fairway and bring it back as close as he dares to the hazard. It just seems too uncomfortable to ask a player to aim out over a hazard to get it back in the fairway, especially if its water. (Road hole excepted here)
Of course, this is just one architects opinion, and I could be wrong. Especially here.
But it is a philosophy of making targets possible to attain, and punishing a misplay, not creating targets that no reasonable shot can attain. It also considers the typed of shots good golfers usually want to hit. There is more, but I have got to go.