I guess I was under the impression that the 1080ti card would be the one GPU clock, and that the DeckLink Quad2 would be more like an integrated SDI converter not a second GPU.
As far as I understand, there is no way to use the GPU as the clock master for the capture cards or other GPUs (unless they have a hardware/software sync feature).
Blending in Arena, the PLS300 doesn't have any blending capabilities just a switch. I would go through two PLS300 switchers because they were available and I wanted to have a reliable bail out option in case Arena were to crash or something weird so just a still on each switcher, each being half of the map.
Yes of course! As a I wrote my response to you, I was working on a show with a similar set but an Analog Way Ascender 32 instead of the PLS300 and we were blending in the Ascender 32 in order to integrate a 2M/E switcher with live cameras in the blend.
Another scenario I have done was for a 3:1 screen and I did the blend on the Panasonic PT-DZ13ks and just used arena to send the blend content, for PiPs outside of the blend region I used the PLS300, for PiPs and graphics in the blended region I did that on Arena,
I typically do all of my GFX work in Arena except for IMAG which I pull from seamless switchers which is why we were blending screens in the Ascender 32.
I never had any sync issues.
I've never had any issues blending projectors when the content came from one Nvidia GPU. My recommendation is to not mix HDMI/DVI outputs with DP outputs so I typically pull my blend only from the DP outputs. I also quite often use the Datapath units for larger blends.