Mounting of separate inlet-outlet part with three bolts makes me guessing, that it might be for future replacement of inlet-outlet terminal block to one that also passes liquid to active cooled backplate like EK. But as it wasn't yet mentioned even though might be important feature, i guess such cooled backplate might be subject of potential future separate product (to not make this block cost too high). Maybe even in some ways of universal design (eg. as block to attach ontop of basic passive backplate) for better recouping R&D/manufacturing costs. For me it would be enough for temps of chips on back of card to be on par on air cooling, as i can max out mem overclock even on air.
And yes, trying to visualize flow path made my head hurt too
, probably due transparency & fusion of block layers in one pic and not exactly seeing where are all the cutouts to pass flow between block layers. When at one point it looked to me that liquid flow over left side of vrm/mem will be stationary, due that inversed-C seemingly connecting together both in and out area over there, i gave up
. It will be much easier to understand from actual block on hands
