I know that, but I don't know what it is that makes a difference to them and I'm interested in understanding it.

I'm with them, and if I had to boil it down to one sentence: Zero mental overhead for representing the idea.

I think visually for anything logic-based - shapes, lines, venn diagrams, etc - so using such abstract no-meaning variables gives me nothing to put into that mental image. So unless the logic concept is familiar enough, I have to invent something for the variable before I can place it in the construct, in order learn the construct.

Once the logic construct is familiar to me, I have no issues with the abstract variables at that point.