I think the point is more that if you’re saying one abstraction is similar to another abstraction, you run the risk of over-analyzing the abstraction level and not the thing-in-itself.