That particular recovery does take a bit of "thinking" to follow the logic in it, but it does work very well. I've never done it with a Deuce winch, but I have with civvy towing equipment that the cowboys at work overestimated the capabilitys of. If you think that looks bad, try it with a short winch cable and four extension cables of slightly varying lengths. That'll keep you hopping for a bit. It does burn up cable fast.
The easiest way to "visualize" how it works is not to try to "see" it dynamically at first, but rather move through it one step at a time. Just picture the cable being strung up and tight-
Then, move the truck backwards one inch... (The front cable won't allow this, but overlook that just for the moment...)
In the picture that Cranetruck posted, start at the left rear shackle of the truck. There is one inch of free cable in that line because the truck moved back one inch. Pull that one inch around the first sheave wheel at the double block. Now on the next line, there is that extra inch of cable you just pulled through, plus one additional inch from the truck's one inch of rearward movement. Pull that two inches of cable around the sheave at the pintle. On the next line, you have those two inches of cable, plus the one more inch of free cable due to that one inch movement of the truck towards the double sheave. Now from that one inch of truck movement and the two you "moved along the system" you have three inches of free cable headed around the second sheave at the double block towards the sheave that is anchored at the side of the truck. There is no movement between these two blocks, and likewise no movement of the sheave anchored at the front of the truck, so that three inches of cable will simply be pulled right around all three sheaves, and be directed back towards the winch. So you have three inches of cable coming out of the sheave in front of the truck, towards the winch. So now, you have three inches of "extra" cable in the line between the winch and the sheave at the front of the truck. The truck has moved rearwards one inch, requiring additional cable in that "leg" of the rigging. (That is the additional inch that was disregarded at the beginning so that this would work "one step at a time" instead of all at once). With that inch now accounted for and removed from the total, there are still two inches of cable remaining in that secton of line between the winch and the front sheave. As the winch spools in, those two inches are wrapped onto the winch drum. Two inches of cable on the drum for every inch of rearward movement. I think that's the hard part for people to see, is that while the winch is realing in cable, it's coming out of the front sheave three times as fast as what the truck is moving backwards.
In simplified theory, that means that this gives you not only rearward pull, but at a two to one advantage. Realistically due to frictional losses at the sheaves, that's not going to happen. In lieu of determining which snatch blocks are being used, how they're constructed, how they're maintained, what cable is being used and all that fun stuff, a good "rough guess" as to the recovery power of this method is that the force available to unstick the truck will be about the same as a single line pull directly to where the front sheave is anchored. Speed will be halved, as that always goes with the "theoretical" mechanical advantage,
As for stretching the truck out... it depends how stuck you are. "Action" members like a winch only exert as many pounds of force as what is available from the "reactioin" member at the other end of the line.
Let's say 10,000 pounds of force is exerted on the cable at the winch drum. That's 10,000 pounds pulling against the truck frame in a forwards direction. At the front sheave, 10,000 in and likewise, 10,000 out on the same line. (Frictional losses aside for now). For argument's sake, let's say there's a twenty degree difference between the two parts at that block. That's 19,700 pounds applied to the block hook and anchor chain/strap/tree, For all intents and purposes, the angle is acute enough that it may be considered "straight". The block to the side of the truck is another story. The angle there is more like (I'm estimating again) 140 degrees. That means the block's hook only sees a load of roughly 6800 pounds. (120 degrees is where the two lines and the hook all see an equal load). At the rear of the truck, in the double block, there is a slight angle of two of the lines, however just like the front, it is very acute and in this case negligable, and should be considered "straight". 10,000 pounds on the line into the double block, 10,000 pounds on the line to the pintle mounted shieve, 10,000 pounds on the line from the pintle mounted shieve to the double block, and 10,000 pounds from the double block to the left shackle. Added up, that's 30,000 pounds applied to the rear of the truck, and 40,000 pounds applied to the anchor for the double sheave block.
With 30,000 pounds applied to the "rear bumper" and 10,000 pounds applied to the front bumper but in the wrong direction, the net result is 20,000 pounds of pull in the correct direction.
Frictional losses abound in this system... While theoretically the pull will be 20K pounds, realistically I'd expect ten. Twelve to 15 is probably attainable with reasonable equipment, but I wouldn't count on it.