I think there are plenty of cases where people prefer not sit on in front of their computer today over going to the fitness studio while preferring going to the fitness studio to sitting on in front of their computer tomorrow.

Changing the near to a far frame changes references. I know that not an exact example of the Dutch Book but it illustrates the principle that framing matters.

I don't think it's hard to get people in a laboratory and offer them different food choices to get a case where a person prefers A for B, C for A and B for C.

I think it's difficult to find general model real life cases because we don't use the idea as a phenomenal primitive and therefore don't perceive those situations as general situations where people act normal but see those situation as exception where people are being weird.