Its easy to poke holes in any applied methodology. I could suggest the time of day wasn't consistent across groups, the seating position might introduce bias etc. But you have to start an investigation somewhere, and I think their methodology was a reasonable balance of practicality.

The work is tantalising evidence for something else going on. Which certainly warrents follow up work. Even if it was just the people were bored and not paying attention to ads, it still has consequences for the cinema business model.