It's interesting to see how some people's brain may prioritize certain effect when it comes to judging a whole image's quality. But after all we're all genetically different right? For me I take many things into account not just one or two features. The best lighting in the end just like good AA or dense geometry, is only a tool to help to make the whole frame look good, it's not a savior of all, far from it. When you get through the initial phase of awe on a macro scale such as a world lit by good lighting, you subsequently start to examine minute details, the quality of assets, resolution, AA etc. But what happens when most of them are rendered sub par or a severe lack of? Your brain would start to depreciate the initial shock and awe slowly but inevitably. In the end your brain would come to a rough conclusion after weigh in on the pros and cons that the whole thing just isn't as astounding as before but feeling deceived. It's not that people don't appreciate QB's lighting prowess, but rather the rendering engine is held back by too many other issues.
To me QB could have a full on raytrace engine working yet still looking slightly above average due to low quality assets, low res buffers and low ish geometry density.