To be fair, their control variables treat the first objection (wealth), not the second (brand preference; and yeah there's some correlation but one doesn't imply the other)