The point of peer review is to stop bad research, not fake research. The reviewers are supposed to make sure that the methods used to acquire the data are clearly documented and that the conclusions drawn from the data follow from the data that was collected.

They can't verify that the data collected actually was collected without literally watching over the shoulder of the researcher while the researcher was collecting it, which is obviously not feasible. A researcher actively trying to deceive the reviewer with fake data but with a good method for collecting that data and analysis of the data that would be correct if the data were real is not something that should be caught by a reviewer. That should be caught by someone reproducing the research after publication.