I'll just toss in a reply for you to reply to and get more posts but...

a resting GPU temp lower than the motherboard/ambient temp, swoon!

D'oh

good improvement from 56 -57 at 1331 and it seems that a push/pull doesn't offer any advantage for your configuration

Your temps are a little high. My Hybrid idles at 25-26C, ambient temp is 72F / 22C. Under load the hottest I've seen is 50-52 while benching, gaming usually 46-48. And this is with the stock fan. You should consider replacing the thermal paste and see if that helps. Or give EVGA a call to possibly RMA, as you shouldn't have to replace the thermal paste in the first place. 

If you are reading the review and trying to compare to your system, then your system has to be the same case and air flow setup to be an apples to apples comparison.
 
Just because it has a fan on a radiator does not mean that you don't need fresh case intake air and airflow.
 
A few tips:
(1) Idle temperatures do not mean anything.  Load temperature and ambient are a much better metric.  What you are looking for is the "delta T", or the difference between load temperature and the ambient temperature (i.e. the temperature of the air the case is bringing in).  In any sort of experiments you do, you should use the delta T as a figure of merit...lower is better.  Air type coolers take about 5 minutes to hit "soak" temperature.  Water type coolers take about 30 minutes to hit "soak" temperature.  They usually take about half that to cool down.
 
... You should check the review to make sure that they were not quoting delta T versus actual GPU temperature.  Many of the review sites are using delta T now for AIO performance...as actual GPU temperature is dependent upon ambient temperature.  If they are not quoting delta T, they will usually tell you that they have a controlled ambient of xx.x C or something like that.  You can compute delta T from that.
 
(2) You are running your CPU cooler as intake.  So, you are bringing heated air into your GPU radiator.  As the CPU gets put under load, the heated air exhaust from the CPU is being sucked into the radiator on the GPU.  So, you are bringing air into your GPU radiator that is already heated by the CPU and the other components in the system.
 
(3) You can't just read the specs on a fan, you have to look at it's P-Q curve, and how well the fan performs at what your system air resistance is.  The Y axis is the pressure for the fan, and the X axis is the CFM for the fan.  A fan produces its maximum pressure at 0 CFM of air flow, and it's lowest pressure (i.e. 0) at it's maximum air flow.  The line between maximum pressure and maximum airflow is not a linear diagonal line.  This is what separates the "men from the boys" in terms of fan performance.  It does usually scale up and down with RPM though. 
 
...There is a European website (Italy I think) that actually measures the P-Q cure for a fan when doing reviews.  They have a "translate" button that will give the review in English.  Their reviews are spot-on for every fan I have looked at.
 
 
I've been around this a lot myself.  I have found that using any of the auto settings on Corsair link (i.e. quiet, balanced, performance, etc.) is pretty pointless...because as soon as the coolant temperature hits 38 C or so (don't remember the number), the fans spin up to 100%.  So, your best bet is to set these at a constant PWM percent.  As the AIO radiator is your main case air intake, you should be running these pretty high.
 
I've played with a lot of different configurations, and have found that the AIO as exhaust air is the "best of the worst".  You don't really need push pull on the AIO radiator...unless you are trying for super silent.  An option would be:
 
- use 2 of the 140mm Noctuas as intake
- put 2 of the 140 mm Noctuas on the AIO radiator as push for exhaust
- Crank up the intake fans to as loud as you can stand it
- set the AIO fans to a fixed % (as loud as you can stand it)
- keep the ML120 on the Hybrid radiator (I would bet that you only need 1, not 2 in push/pull...up to you)
- set the Hybrid fans to a fixed % (as loud as you can stand it.)
 
Now, start playing around with different fan speeds for the AIO exhaust.  You have 2 140 mm fans sucking out air...and if you don't have enough input air, you can starve the 120 mm fan and get less air flow through the AIO radiator.  If you are trying to minimize your GPU delta T, then you need to make sure you have the best airflow possible through it's radiator.
 
I stated in an earlier thread that a 120 mm radiator for a 250 W GPU is to low.  Heck, you are running a 280 mm radiator for a 150 W CPU!
 
 
In my setup, I am running:
 
5820K overclock at 4.3 GHz (Corsair H110i GTX AIO...280 mm radiator)
32 GB memory @ 2800 MHz
Titan X Pascal (on air)
EVGA 980 Ti Hybrid (stock fan)
 
My intake fans are 2 ML140s (red of course - hehe).  I keep these at 75% RPM.
I have 2 ML140s as push through the AIO radiator as exhaust on top.  I keep these at 55% RPM.
 
I actually monitor the coolant temperature on the Corsair AIO versus the CPU temperature.  (The CPU temperature is a function of the coolant temperature and processor load at steady state.)  When I am not gaming, I run this machine 24/7 for [link=mailto:folding@home]folding@home[/link].  My typical ambient on my office is about 28 C (I have other PCs on my office).  At full load on both GPUs (both overclocked) I get:
 
Titan X P: 77 C (49 C delta T...this thing needs an AIO!)
980 Ti Hybrid: 52 C (24 C delta T)
CPU AIO coolant temperature: 36 C
CPU Temperature (while folding): 50 C (22 C delta T)
 
 
I hope this makes sense!

He's at 56 C GPU, with 25 ambient...so 31 C delta T.
 
Considering he's bringing hot air into the case through his CPU cooler...I don't think his temps are out of line.

So EVGA support say my temps are a bit high for a Hybrid, but don't seem concerned saying that cards that idle in the 60C's can still only go to 70C under full load. So what exactly is the benefit of paying extra for a Hybrid card, because it sure doesn't sound like running cooler or quieter is guaranteed. I'm wondering what this FTW card would be doing under air cooling only.

No worries, I don't think you're being a jerk, I can't correlate the results either which is leading me to think something is wrong/faulty.
 
I agree with a lot of your points, especially regarding the intake through the CPU radiator when the CPU is loaded, it's why I decided on keeping those fans as low as I could and using the other intake fan to do the heavy lifting of case cooling. I know that is working because my case is always maintaining positive pressure and I can feel the cooler air ejected out of the rear vented openings.
 
My MB temps are totally in line with other reviews and my expectations for this board inside a case, and I'm actually seeing better MB temps than hardware canucks saw on their open bench test with 21C ambient.
 
My idle temps are taken 15-20 minutes after first turning on, and at least 30 minutes after any load testing to ensure everything is stable. My room does adjust slightly (less than 2-3F) during load testing and takes about 5-10 minutes with the AC to come back down afterwards. I have a digital thermometer sitting on top the case the measure ambient.
 
I'm going to see if I can borrow the IR thermometer we have here in work, I'll use it to measure the case intake and exhaust temps as well as the Hybrid radiator.

That's a shame.
 
Good luck with the RMA!