Update on the whole topic:
So I now have confirmed from multiple 1950x users, that my system should pull closer to half of what I´m seeing with exception of idle. It seems like the higher my power draw gets, the less realistic it is. Like 200W idle is maybe off by 10%, but the 1200W in gaming load should much rather be 600-700W at best.
I took a deep dive into what I remember from classes about electricity and how it works and my current theory are high non-linear loads from the PC / PSU, causing some severe overtones on the third harmonic which then get picked up by the wattmeter since it has neutral wire pass-through.
To explain that for everyone in short: Power delivery is split into 3 seperate phases, since this is what most electric generators (motors) work with (3-Pole) and by doing so it reduces losses from source to drain. The protective neutral wire that is used for equipotential bonding in theory carries no current because the 3 phases all have a shifted amplitude by 120° and cancel each other out on the neutral wire. (remember that AC is delivered in a switching freq.) By applying a non linear and fluctuating load however (PC-PSUs being a prime example), you destroy that balance on the neutral wire and even encourage the effect to turn from canceling to actually being additional. So the N-wire can carry multiple times the current of a single phase load which in old installations can cause fire, because the wire gauge used to be small since it usually carries no load.
Long story short, a PSU with such a high capacity has huge capacitors with an extreme inrush current on turn on, and the very alternating loads that a pc needs depending on what it is doing can cause these overtones which result in a current on the neutral wire. And I think that exactly this is happening and my watt meter picks that up.
No idea if that is true or not, I´m not an expert on any of this. Just my thoughts.
My additional steps so far:
Tested a different PSU from Corsair (RM850i) and it showed a lot less power draw under load. Idle was more or less the same (~180W) but the CPU load test had it at 325W and the full load with GPU settled around 750W. Which to me shows that my initial feeling was correct, the numbers with the EVGA T2 are way higher then they should be.
Now that I already have it open and a second 1600W T2 that I bought new for another project, I will swap the new one in and see if it behaves any different. I don´t expect it too, I think it has something to do with the high capacity, peak loads and the way my power meter works.
If my theory is correct then there is no worry for my PC parts, it all works fine. Just important to note that this PSU can set your house installation on fire if you have an old one with thin N-wires!
