You've put your finger on something real, and it's not a bug we're trying to hide. It's close to the whole point. The tool doesn't define "mind's eye" or "visualization" on purpose. Those words mean wildly different things to different people depending on your inner experience, and that ambiguity is exactly what we're trying to surface. When we line up what someone generates against their VVIQ score, the interesting result isn't "here's an accurate picture of your imagery", it's how often two people with the same score produce completely different images, or describe the same experience in incompatible ways. The divergence you're describing is the signal, not noise that makes the answers meaningless. That said, your specific point lands: if you have aphantasia, no generated image is going to correspond to what you "see," because there's nothing there to render. An image generator can't draw absence, so many people depict it is black. You're pointing to the fact that, by definition, you can't visualize something that has no visual quality. However for those who do visualize in their mind's eye, they are able to get the sliders to something that represents their inner experience. We often hear there's not enough variety - ie they want the horse to be moving or look more like a cartoon.