First of all, fingerprints are accepted as the gold standard of physical evidence, but they never were subjected to the scientific validation required by the "Daubert Test" or "Daubert Standard" for scientific evidence. Moreover, rather than actually matching a full set of good prints against a full set of known prints, "experts" testify on the basis of partial prints and latent prints. Latent prints are copies, sometimes copies of copies. Moreover, fingerprint experts are not allowed to disagree with each other. Either they agree or the results are "inconclusive." No one is allowed to be wrong at this. Fingerprints have been faked by criminals; and the first known case of a faked fingerprint (1925) was revealed by the FBI to have been perpetrated by a police officer. (See "Suspect Identites: A History of Fingerprinting and Criminal Identification" by Simon Cole (Cambridge: Harvard University Press, 2001).

From there, junk science devolves into psychological profiling, dream interpretation, hypnosis and regression, shoe prints, tire tracks, fiber analysis, and rifling marks on bullets. Each or any of them could be an incidental or circumstantial fact to corroborate other evidence, but not one of them is scientific.

DNA is, indeed, scientific. It meets the Daubert Test, for expert acceptance, peer review and testing, and double blind statistical sampling, and a standard of falsifiability. However, even DNA evidence has been compromised. Bad storage, mislabeling, and other errors do happen.

Writing about the so-called "CSI Effect" Gregg Barack, Young Kim, and Donald Shelton focused on jurors. Among the many challenging discoveries of their statistically valid investigation was the fact that less-educated people demand more physical evidence. I point out that those people are least capable of judging it. They are what Ayn Rand called "muscle mystics."

See
“Examining the ‘CSI-effect’ in the cases of circumstantial evidence and eyewitness testimony: Multivariate and path analyses,” Young S. Kim, Gregg Barak, Donald E. Shelton; Journal of Criminal Justice 37 (2009) 452–460
"The 'CSI Effect': Does It Really Exist?" Donald E. Shelton, National Institute of Justice Journal 259. (17 March 2008) http://www.ojp.usdoj.gov/nij/journals/25....
“A Study of Juror Expectations and Demands Concerning Scientific Evidence: Does a 'CSI Effect' Exist?” Donald E. Shelton, Young S. Kim, Gregg Barak, Vanderbilt Journal of Entertainment and Technology Law Vol 9 No 2

Since the declaration of scientific criminology in the 19th century, police and prosecutors have sought out empirical evidence, relying on what were then new sciences such as chemistry. Cesare Lombroso claimed that scientific measurements could identify congenital criminals. Dactylography (fingerprinting), graphology (handwriting analysis), polygraphs (lie detectors), and psychological profiling, were joined by laboratory analysis of fibers, hair, tissue, cloth, paper, ink, tire treads, shoe prints, typewriter keys, and just about everything else.

Sir Arthur Conan Doyle's Sherlock Holmes in the Strand Magazine became the public's model detective. Not a mere rationalist Holmes's theories began with empirical evidence. He studied cigar ashes. In his day, the electron was a theory. Fifty years later high school students accepted it as basic knowledge. Today, surgeons are guided by images created by positrons, the anti-matter analog of the electron. Humans have been to the Moon; our cellphones depend on satellites. We clone animals and genetically modify vegetables. It is not surprising that through this century, empirical evidence from the police laboratory is expected to reveal and convict perpetrators. But not all science fiction becomes science fact.