You may want to upload a few pictures. Mid to high 50’s is relatively high, but only you can see the block, so no one can see why to block isn’t filling properly or if it is sitting properly. I suggest pictures around the edge of the block/card, as well as pictures of the front of the block as well.

Well the potential issues i suspect are:
- block not seated properly, due to maybe a wrong length screw or not tightening them around the core first (and them undoing them to fit longer ones where the backplate uses them, no mention of backplate in the block manual, you only find that out in the backplate manual)
- thermal paste coverage not that great, as i had to correct the block placement a little to get the holes to line up. I used the X and + pattern, not the blob that was in the manual.
- air in the system or in the block, haven't titled the case with the reservoir open yet. The micro bubbles are gone now from most of the system, except for some in the gpu block.
- having the rad fans curve based on the CPU temp, so the GPU just gets the water it gets (but that might be only relevant for GPU stress tests, where there's minimal CPU load). Haven't raised fan speeds or pump speed after adding the GPU block to the loop.
 
Here's the pic of the not fully filled block, after 5 hours of 100% pump and around 15 hours of normal usage (40%) it still hasn't filled any more:
http://shrani.si/f/3h/13f/3wGnl0R8/img3443.jpg

Full setup (not very EK friendly this TT case)
http://shrani.si/f/I/Em/1JiUbduY/519607872165225903807894.jpg

Will add the closeup ones of the block on the card a bit later.
 
I leak tested with only the pump on PSU power for around 3 hours, titled the case but not to extremes, there was some air burping but the GPU block water level hasn't changed. I maybe filled the reservoir a bit too much to be able to tilt the case with the cap open to any extremes though. Idle sits at around +4-5 above ambient, which is probably good, around the same as CPU idle that is 24-25 (at the moment 20°C room temp, CPU is at 24, GPU is at 25°C). Would idle be like this if there were thermal paste or contact issues?

Feed your pump from the res first

outlawii
Feed your pump from the res first

+1
 
Return line to reservoir, reservoir to pump input
 
Those pictures really help us understand your current setup
 
Your radiator could be used as a reservoir if the fittings were at the bottom & connected to supply side of pump - on top you could be sucking air as that is were the air sits
 
Running the pump dry causes damage as the liquid both lubricates & cools the pump

Worth noting those EK CPU block have an assigned In & Out - they only flow properly 1 way
 
I can not tell from you photo if the pressure side goes to the In port on your CPU

Ok, it just occurred to me that i can put a combo reservoir on my pump, so i ordered one along with some 45° and 90° angled fittings to help with bends, this is how it will look then, should be good?
 
Basically i'll shorten the current CPU to reservoir line and fit it to the GPU intake and connect the GPU out one to the other port on the new reservoir, as both will be next to each other on the side. Maybe i'll need to swap them around as they have assigned in/out, if the on closest to the fans won't be the out one :)
 
This should be good then until 7nm RTX comes out, hopefully Fall 2020 If there are any additional ports on the side of the new reservoir, i'd maybe use that as a drain point, about time i have one
 
 
 
 
 

Cool GTX
I'll look for your next update when the new parts get installed

I've spent 6 hours yesterday fitting the reservoir to the pump and changing all the hoses except one, most of the time i couldn't decide between using both ports on one side on the pump or one on each side, decided for the latter in the end. Also the CPU to GPU bend was a little problematic. When filling the system up, the card again wouldn't fill up (what's funny is when i drained the old setup's reservoir, just spilled it like from a glass, water came up from the GPU block and it was full ). This time i tilted the case in all possible direction the PSU cable and reservoir water level permitted without the pump running dry and i did manage to get a lot of air out, but still the left side of the card is better but another step like that would fill it up completely, the right side is now full. I've also flattened the pump curve to 35% or 2950rpm up to 70°C, if maybe 4000rpm that was average before would be too much flow, not giving the water enough time to cool off (the pump is rated up to 1000L/h, my loop is supposed to need only around 300L/h)?
 

 

 
Haven't played any games yet, but a FurMark stress test at 20°C room temp before started at 26°C and ended at 57°C after 20 mins, now it went from 25°C to 53°C, so it's an improvement. I expect the average gaming temp to drop now from 51-53°C as well, should stay below 50 now.
 

Cool GTX
that looks like an Air lock, to me
 
As air will always rise to the highest point, you will need to move your PC almost 90 to get that out.
 
The fluid can compress the air, but without a large flow based on the design (hump) it is unlikely to be completely removed without tipping so the air can escape.

If need be, you may have to unbolt your pump to keep it aligned while moving the case, so you do not cavitate the pump

 
Spent a couple of hours today tilting the case in all possible directions with the pump at 100%, finally managed to clear that big air bubble almost completely. No change in temps though, not much improvement compared to the previous setup, gaming average temps are still around 49-51°C on stock clocks.
This has got to be either poor thermal paste/block contact or just the max a single rad setup in an open case can do. I did lower the pump speed from average 3900rpm to a fixed 3000rpm. Would probably need a temp sensor reporting water temp to the motherboard and have the fans work based on that, not the CPU temp.