If reductionism is true, then Holmes is real*
* Holmes would exist as a particular brain state of Arthur Conan Doyle.
I'm pretty sure that that's not how Holmes is defined. But the idea or concept of Holmes is a real idea or concept all right.
I don't see what the problem is there. Distinguishing between the reality of an entity as a real thing, as a concept, and as a phenomenon in the brain seems to me well within what an average human should be able to understand, unless they managed to get stuck at some point during their childhood mental development stages (as per Piaget.)
You probably can distinguish between, say, the concept of "cat" and an actual cat, or between the concept of justice and an actual lawsuit, or between SF concepts like flying cars and lightsabers and actual working models of either. Ditto for their existence. The idea or concept of a flying car exists since at least 1905 (or millennia back if you count other kinds of flying vehicles, like Ra's boat or other gods' flying chariots), but an actual commercial implementation does not exist. The concept of a lightsaber exists since at least 1977, an actual lightsaber does not exist. Etc.
There's no particular need for reductionism to understand that. One just needs to be capable of basic abstract thinking.
All that reductionism adds there is a mapping to a brain state, but no real paradox. The same continues to apply. The idea of Sherlock Holmes exists, the brain state for it exists, an actual Sherlock Holmes does not.
