Utility is ordinal not out of some deep desire on the part of economists to "appear scientific" but because it reflects a simple empirical fact about humans. "Utility" is a measure of value to someone, and human values are ordinal: I value my children more than my cats, and no number of cats will ever equal the value of my children. It would be incoherent and wrong to say I value my children ten times more than my cats, and no cardinal relationship between them could possibly represent reality.
This is not mysterious or esoteric. The ordinality of human values is what makes rational choice under uncertainty such a difficult problem, because we can't compute expected values from ordinals (unless there happens to be a market in the things we value, in which case we can get cardinality through prices... but there is no market in children or many other things we value.)
By making it sound like economist's interest in publicly testing ideas by systematic observation, controlled experiment and Bayesian inference--which is what the discipline of science actually is--is some kind of inexplicable fetish instead of a perfectly reasonable goal for anyone who wants to say anything meaningful or interesting about the world, the author does everyone a great disservice.