> The N-back task is useful because the degree of difficulty can be systematically varied. It requires a fairly naturalistic combination of memory and attention. This is to say it is a good test of our ability to think hard:

That's not what I would consider thinking hard. Usually that would be building up a large mental model of many things and complex relationships, then think about different ways to re-structure it and/or imagine mutating data passing through such a structure.

That experiment setup is something I would not enjoy, and may just take the zap as the task itself is meaningless.

A better experiment would be something that rewards thinking hard--something like a card deck-builder game that can be won for bragging rights.