I can’t see the light in your example either. If I shut my eyes, and try to see it, I just see blackness though I have no problems with the concept of a blue light (or even the entire scene) - I just don’t
see it. To answer the photo question, I don’t have a verbal definition of a photo, I have a relational one. I can still describe the photo reasonably well, but I don’t “see” anything.
I’ve thought quite a bit on how my mind works once I discovered that “the mind’s eye” wasn’t just a euphemism for other people. I think my mind works with rules rather than visuals, and it constructs/deconstructs a lot of them as needed to describe something it can work with.
I’m a software engineer, an ICT5, working at Apple. I tell people my primary skill is “solving problems”, and that’s accurate as a 10k-foot view. I’m good at mentally modeling the intricacies of how a system works, what the dependencies are, what the interactions are between sub-parts, how that effects corner-cases. I can “run” this mental simulation in my head forwards and backwards a bit (not too far) and get a grasp of how it fits together. But there’s no visual to any of this, whether I’m thinking about a bicycle or server-side cryptographic mail storage and transport. It’s all just interacting rules. I’m often the one to point out a flaw in a systems model well before we get to implementation, and I’m good at “breaking new ground” when it comes to implementing something, even for a different group - I get “loaned out” a fair bit.
So I think there are upsides to not having a mind’s eye - I chose physics at college, and found it easier than most I think; equally there are upsides to having the ability to visualise things, I expect. It’s probably a species-optimization to have the capacity for both…