I would characterize GDP as "obviously wrong" as a measure of actual productivity.

That being said, one good reason to assume that someone is underpaid relative to their output is that they work in the nonprofit sector, where salaries are typically lower for equivalent levels of work.