One thing that may have contributed to the perceived reduction in randomness is the RTJ style (and Wilson, Lee) of flanking bunkers on each side of a perceived landing zone.
I believe, Ross, Colt and others mixed in a few as targets (usually behind the landing zone) and many more as angled carry bunkers, all for the approximate same tee shot. I believe Ross used 200 yard dogleg points, whereas RTJ moved them to 250. Ross may have planned for 225 shots later in his career, I don't really recall.
Add it the idea of staggering bunkers like far left and shorter/carry right, and bunkering becomes far less monotonous.
I think there is a balance to calculating out bunker locations. Say I play on a 290 drive (about PGA Tour Average) with 275 carry. I factor in wind and elevation changes which alters that, sometimes by 10% or more. But, if I find a nice knoll at 259 or 306, etc., I easily put the bunker there, because almost no one hits it whatever the average distance for their "class of player." I also figure that a bunker can catch a flying shot or rolling shot, which makes any exact distance placement not so critical.
As to forcing them in, it rarely seems to work. You need a nice upslope, or at least level ground. If the area happens to be on a downslope, you find it hard to put a bunker there.
I recall playing Bear Creek when I moved to DFW, a Ted Robinson design. He had a few fairway bunkers placed at the dog leg on the downslope, and the backing berms had over 20 foot of fill in them. I really don't need a specific bunker to use that much fill, although I will say, in outdoor scale, its not quite as drastic as it sounds. Also, on some hilly sites, you might face half the fairways in a situation like that and decide that once or twice you will invest the fill to have some bunkering variety where otherwise there would be few fairway bunkers to suit your taste.
Really, that is what architecture is about - literally hundreds of decisions balancing the play, the land forms, etc.