While that is true and understandable, I do with the "what settings can I run with my old card questions" had better answers.All cards get tested with the same settings, it's called having controlled variables and is the correct way to produce comparative results.
Almost no tech reviewers ever want to try to see what settings would actually work. A crude "Crank it up to max and run it" is what they do, whereas I know that certan effects (mainly bloom, motion blur, or depth of fields) get turned off by me ASAP but nobody ever guides me on what kind of performance I can expect with that stuff off. Or seldom, I think notebookcheck do test high as well as ultra.