I don't think you understood the respondence I had and why I had that response in particular. Earlier in the thread it was stated that the difference between a Maxq and full 1070 was about 7% using 3dmark. My response was that in certain testing, especially in synthetic loads, results can actually seem much closer together than they are in actuality, and the use of a full-game benchmark would testify that the difference between cards can be anywhere from 20-40% in actual usage scenarios because the load that 3dmark puts on the GPU isn't the same as a load that far cry 5 or other games would. In fact, I think you touch upon that yourself in this post. 4k 3dmark ultra is meant to strain a GPU, using textures, shaders, and effects that wouldn't need or find a use in real world scenarios.