RL Keeping me busy here, so slow to reply.

I try to stay open minded that is for sure, I have seen a lot in this business and bet I will see lots more before I die.

"Other ATF Specs" - your right, I did not post those spec's or the testing procedures for ATF on it - it is much more complicated than the simple Timken bearing test; great test for bearings alone, but don't fly on most automotive stuff as there is more involved.
I do know that oils are made by these test specs, and 2nd grade oils are normally sold by the 'other guys' (store brands mentioned) not the big guy's.

I think we will disagree with each other on viscosity; the report you posted even agrees with me on that as it quotes over and over how viscosity performance (over a temperature range)was the main thing they were looking for. (Pages: 3 / 5 "Testing methods: Viscosity Retention - etc) - other things mentioned within.

Performance between different brands is a given variable - I have seen "Valvoline" work fine, where "xxxxx" wouldn't work at all. It all comes down to quality control, and standards of the company. There are companies with high standards of control and those that say "Look its red, must be ATF."

A bit in detail:
Viscosity is the main issue when dealing with an automatic transmission. Fluid dynamics inside the torque converter change based on the viscosity of the fluid over a range of temperatures. If the fluid is too thick, it drags on the stator and if it is thick enough the car won't even start (have seen that) // by the same token if the fluid is too thin less kinetic energy (hydraulic energy) is transfered to(from) the stator so the "stall" goes up (sometimes WAY up) - we have used jack or power steering fluid to raise the stall in some roundies.
The physics behind this is noted as: r N5 D2 -- where R is the viscosity (called weight) of the fluid,
http://en.wikipedia.org/wiki/Torque_converter

To put this simplistic: If you take 100 square feet of 'air' and compress it down to about 1.5 cubic feet you end up with about 1500 PSI of pressure. If you took 100 sq feet of ATF and compressed it down to 1.5 cubic feet; well you would need metal designs from Star Trek to hold the pressure - because fluids do not really compress. Where if you took a 'thinner' ATF it would not be as much pressure - try it with water, then ATF just for giggles.

Or try this simple test.
Take any standard transmission (normally 90w oil) and put in 140w "hypoid" rear end oil. Drive it in summer, then try to shift it cold in winter (I have seen people break the shifter trying)

This brings us to "Sheer" / what the Chrysler report you quoted said, what I tried to avoid to some degree / and what the Timken test starts to look at. Please note that ATF Tests are far more excessive than the Timken test.
Sheer is where the fluid breaks apart (at a molecular level) because of the pressure it endures and or temperature (to be fair). This is where the additives, mentioned in that report, come in. When the fluid breaks down, it looses viscosity performance because it is now two separate fluids again (although mixed).
Again, viscosity - this time over a specific period of time.

Lubrication is the next point I think you made, and a good one as it shows how ATF has three functions (rust prohibiting being the third - not discussed). Not only must ATF retain its viscosity, it must provide for lubrication as well (modern chemical engineering). Again however we are dealing with quality control between the brands.

The last point I will make before telling two short stories is that these issues are very well understood, and an attempt at making a 100,000 mile fluid is what ATF+4 is all about (and it is NOT suggested you try to run it that long - be smart). I will admit that I know nothing about mining equipment, or heavy equipment either. I worked in different auto shops over the last 35+ years and built a "lot" of 1/4 1/8th and roundies.

Story 1: I don't remember the car, but it called for something other than DexIII - but that is what we put in it. Now, this was valvoline as we got a deal and had a pallet of it in the back. Come a few months later the same basic car came in again, and we used non-valvoline DexIII in it. It would not shift, would slam into gear, if it did shift, and the lockup never did. We had to flush it and put valvoline DexIII back in it - no problems.
So different problems can be explained by the 'brand' as much as the type.

Story 2:
My reason for coming here in the first place was that I have a used 99 46RE (Ram truck). It had a shift problem so I changed the fluid and filter, went back with ATF+4 (some lower brand name; was on sale /shrug). In any case the thing would shift 'light' when cold - and not at all when hot (less than 2 miles driving). It would also flutter or seek on shift during the mid temp's. I also read that there is a problematic check ball near the radiator so decided to put on an external cooler & flush the system while I was at it.
So I replaced all the fluid with DexIII Maxlife, and test drove it. Now the shifts were very firm, and no 'shutter' on TC lockup (was firm there too) - downshifts pulled power no problem. The problem returned on longer trips (15+ miles) - cooler and thicker fluid simply extended the drive time.
At this point I called a trans friend of mine and he said it was the gov sol - so I pulled it (screens were full of trash (from the flush?)) in any case I cleaned it and put it back together with 1/2 DexIII and 1/2 B&M Trick Shift. Now you can't even feel it shift, unless you hit the gas then it shifts like the DexIII (Firm).
Hindsight being what it is I should have just dumped the lower quality fluid, replaced it with a high quality brand (still with a touch of B&M) and I bet I would have been fine.
Moral to this story: Brand name does matter.

For me all the fluids are the same (exception of Dexron VI) the thickness changes the shift and that is about it - same reason you used to change the spring tension on the valve body to change the shift point (shift kits). Price / Brand, mean higher quality control - more to specs.
IMHO the report you linked confirms that, but that may not mean we will agree