my current understanding...
-the tubule loop - you go down into the medulla and it's permeable to water. the tubule filtrate gets more concentrated as it moves deeper, matching the surrounding osm. then as you turn to go back up, you start letting ions out while not letting water out. you're moving upward, to less concentrated tissue, and letting out ions makes the filtrate less concentrated too. then partway up you start actively pushing out ions. most loop images show 300 at start, 100 at end, and this is what makes sense to me because the point of the loop as i understand it is to concentrate the medulla - to have the filtrate actually leave some salt behind and thus be hypotonic when it comes out of the loop. the point of this loop is to build the gradient in the medulla.
the loop of henle countercurrent idea seems to be explained well by this
http://www.colorado.edu/intphys/Class/IPHY3430-200/countercurrent_ct.swf
note that when they start with everything at 300, at any horizontal level you're concentrating the medulla relative to the ascending tubule, which causes the descending tubule at that level to be concentrated by passive water diffusion out. when that filtrate moves around to the ascending limb, it is already a bit concentrated, and the salt pumps are now establishing a gradient from a better jumping off point, so the medulla at that level gets more concentrated, therefore so does the descending limb at that level, the filtrate moves around, etc.
-vasa recta - the other countercurrent exchange is in the blood vessels, and is different from the tubular one because salt can move freely in and out of the capillaries of the vasa recta. this means salt moves out into medulla on the way down, and back into the vessel on the way up. it's all free movement so it does not change the medulla's osmolarity. it just prevents washout.
in that it prevents loss of something it is like the heat countercurrent exchange seen in vessels in penguin flippers and the like.
