Ron,  Sorry this has took so long, been tied up.  A few comments. 
Ron Gratz wrote:
JBarca wrote:
Now enter the compound angle turn.  When the hitch head rolled and unloaded the inside turn WD bar to zero or near zero,  there is only the outside turn bar left to provide any level of WD. If the outside bar load stayed the same, then the truck would of lost 1/2 the WD and the front of the truck would of came up enough the driver “may” feel it.   Lighter suspension trucks would fell it more so then stiffer suspension trucks.
For reference, the OP's Excursion rose 0.44" in front and dropped 1.38" in rear with no WD applied.
Using 800# bars, he was able to drop the front 0.31" and raise the rear 0.62".
If the WD bar load were reduced by 50%, the front would go back up about 0.15" and the rear would go back down by about 0.31".
Therefore going from the full WD bar load to 50% bar load would change the slope of the TV by about 0.46" over a distance of 137" corresponding to an angular change of 0.7 degrees.
My point is that reducing the WD bar loading by 50% might not make a noticeable difference in the stance of the TV.
I think if we really want to know how the WD bar loading is affected by different simple or compound angles, we need to come up with a method for directly measuring or calculating the load on a WD bar.   
Something is not totally adding up with the fenders and weights of an 1,100# loaded TW when mrekim’s had his frame issues.  My 2003, K2500 Suburban gasser front end rose  1 1/8” with a 1,200# loaded TW and no WD.   However mrekim has the heavy diesel up front and the fender rise effect may be less for rise even if the weight is moving.   We can let this go for the moment that not all TV’s may notice the shift in 50% loss of WD.   
And I totally agree we need a way to directly measure or calculate the load on the WD bar.  More on the loss of one WD bar to zero later in this response. 
Ron Gratz wrote:
JBarca wrote:
Since the TT TW did not change, it is still providing a constant force down on the only WD bar that can produce tension.   Since the same TW mass is available to only 1 WD bar it comes down to, will the hitch head angle and the rising rear of the WD bar allow an increase in tension?
The tongue weight does not determine how much load is generated by a WD bar. Hold that thought.
The load on the bar is a function of height difference between front and rear of bar, slope of the front end of the unloaded bar, and bar stiffness.
The user determines how much load is generated by changing some or all of those parameters.
If you simultaneously raise the rear of a bar and decrease the front slope (e.g. decrease the head tilt), you can have an increase in bar load, no change in load, or a decrease in load.  
Here I need some help understanding why TW is not a factor. I do not understand how you can discount it.  The hitch was adjusted because of the TT TW. TW is one of the forces pressing down on the WD bars creating tension in them.  The more down force, the more tension and less down force is less tension.  So far so good? Agree? 
I agree that “The load on the bar is a function of height difference between front and rear of bar, slope of the front end of the unloaded bar, and bar stiffness.” are factors in bar tension and so is TW. In this case, the hitch was adjusted for a given TW where the user adjusted the head tilt and or chain length to compensate for a given TW.   Both WD bars once the hitch was adjusted with the TV and TT straight in line on level ground are close to equal tension in the bars along with equal loads on both sides of the A frame from the WD bar chain loads. 
When one WD bar lost tension due to hitch head roll and the tilt is relaxed, the WD on the other side of the hitch head truck can still be under tension and in most cases is.   When the back of the truck drops in height due to loss of 1 entire WD bar going to zero, the angle of the hitch head is tilted back towards the TT by the truck dropping height.   Similar to a back flex when the truck goes up hill just not as pronounced. 
The TW of the TT is still pressing down and more TW mass is available to be resisted by the only  remaining WD bar which has tension.  While the head roll angles play a large role in how much  weight is added to the one remaining WD bar, the head now has more rear tilt then it did before, the rear of the bar is rising up the cam and more mass available to load on the one tensioned bar then when the loads were shared by both bars. 
My point is, TW does play a role in this, it cannot be discounted.    In order to bend the snap up, the bent snap up side of the A frame had to gain more chain tension and the WD bar had to have an increased tension in order for the chain to have more.     I do not think you are disputing the fact the WD bar tension had to rise in order to increase the chain tension.  TW is still a mass regardless of head roll angle and can create more load in only one WD bar pending how the head angles end up.  Granted the head angles may allow one bar to go to zero and the other bar be less, however if the head angles line up, the one tensioned bar now has more downward mass only held by 1 WD bar and the bar tension can go up.  It is a combination of the angles, the TV turn and the mass pressing on the bar to create the tension.
A normal way I have seen both WD bars unload is when the truck goes over steep hill heading downhill while the TT is still on a more level ground or still going uphill. Like the TT going uphill on a steep RR track crossing when the truck is over the hump going down hill.  The bars unload when the TT is going up the hill and then heavily load up when the TT is coming down off the hill. 
Ron Gratz wrote:
JBarca wrote:
In order for the snap up to fail like they do, an increased force presents itself to one side of the WD hitch at the WD chain that overloads a non bolted snapup.   A hypothesis is, that the outside WD bar rose enough in tension from the compound angle turn to cause the snap up failure.  Now the task is, how to quantify that?   If this compound angle turn is not creating the force, then what other combination of actions is?
My current thinking is that the friction force between bar and cam results in an amplification of lift chain tension, and it is that which is causing the problems.
Of course, I always reserve the right to change my mind.  :)
Ron
I agree high friction is in the middle of this.  When the load goes up, the friction goes up and as you have shown, the friction load is large multiplier, possibly 4.3 to 1 if I understood correctly.    However my recollection of the laws of friction remember that to increase the friction force we have to change 1 or both of 2 things. The coefficient  of friction has to change to a larger value (not likely in this case) or the force applied has to increase.  (most likely in this case)  If we have no force increase, how can the friction force increase?   That remains the question,  how did the force increase?….:h
I will see if I can hunt up some kind of load cell for the chain and or the load pointer on the WD bar. However I do not have a way to record the constant travel under tow.  Yet anyway.  Also need a positive way to hold the hitch head still while measuring load applied to the end of the WD bar for the spring constant data.  I'm a think'in.