If it helps (should), here is the Intel industry-standard SATA power component-end pinout you can check with a digital multimeter to be sure it's correct.
 

Anyone?

Already have just waiting for a reply. Guess I'm just impatient! lol

To be perfectly honest I've not got much experience using a digital multimeter. I watched a few video guides on it and actually started with the 4-Pin Molex first as it was easier to test and understand. From what I've read with these GQ units while labeled differently (PERIF1 instead of SATA) the keying and pinout on the back of the PSU is identical for both 4-Pin Molex and SATA. In the case of the 4-Pin molex I was getting some equally strange readings that would change based on which ground I used in the middle. I either got -12v on both pins at the edges of the connector (the ones that should be +12 and +5 respectively) or -5v again on both pins at the edges of the connector if I used the other ground. It was almost as if power was being sent down the ground wires.
 
Then I used a SATA pinout like the one shown above to put the black probe on a ground and the red one on the 3.3v, 5v and 12v. I was not however using the tiny pins on internal part of the connector as they were hard to reach but instead using the pads on the back which groups the 15 pins down to the 5 basic wires. I'm fairly certain of my results as the SATA connector's keying makes it easy to work out which pin is #1 etc. I also ruled out my lack of knowledge and the DMM by doing the same on my Seasonic unit and while I didn't test 4-Pin Molex I did test the SATA and got the expected readings of +3.3, 5v and 12v.
 
So it strikes me that either the PSU is faulty or the cables I have for it are not the correct ones, hence why a rear pinout would be useful.

Indeed, like I say the readouts were fine on a different PSU so either the cables are the wrong ones or the unit is faulty. I was actually using the unit in a different system but only the 24-Pin, 12v 4+4-Pin and PCI-E were connected as the drive was M.2 and didn't have any issues with stability or anything.
 
I was testing the unit outside of a case (using the included 24-Pin power on adapter), so the ground from the pin was the easiest one I had to hand. I could retest it but honestly I fail to see it making any difference and I'd rather not modify the cables.
 
I wonder if EVGA will send me a set of replacement cables instead, still waiting on them to reply but it might be worth asking.

Oh they actually have a UK phone number, guess I'll do that then, thanks. :)