Agreed. Once a file/page reaches a certain size, I find minimaps to be a mostly unintelligible blob. Maybe it'd be solved by just making it bigger, but then we push into the problems of screen real estate and what not.
I think, it's also, instead of just approximating the position in the document, we now feel compelled to translate our mental image to a visual one and to match this with the rendering in the minimap. Which introduces an additional step of dis-abstraction, and what is essentially abstract (text) now has to be treated like a material object, as well. Moreover, if the overall structure isn't as varied as it may be the case with code, it doesn't provide much significance, anyways.
(It may depend on where you're located on the aphantasia to photographic memory continuum, though. I'm definitely closer to the former than to the latter extreme.)
(It may depend on where you're located on the aphantasia to photographic memory continuum, though. I'm definitely closer to the former than to the latter extreme.)