If it cools down to the liquid on the video card using only the wb on the GPU (so it's not a full cover) there is a risk to the ram of deterioration or premature failure (or in general, a shorter average service life) compared to the state of normal receiving by the sink factory?
Cooling the mosfets is more critical than cooling the ram since the VRM will throttle if they get too hot. Most GPU/CLC kits include a fan for this. All GPUs that come with them also have cooling for the ram and VRM.
So it is better to cool the memories and the VRM right?
I could use small heatsinks attached with the thermal conductive double-sided pad (both already in my possession) on the memories and the VRM, placing an 8mm fan that blows towards them, as well as the WB on the GPU.