Incidentally while Scalia's dissent accuses Kennedy's majority opinion of slanting the facts to support his argument the dissent does not appear to be perfect in this regard. In footnote 2 Scalia claims:
By the way, this procedure has nothing to do with exonerating the wrongfully convicted, as the Court soothingly promises. See ante, at 17. The FBI CODIS database includes DNA from unsolved crimes. I know of no indication (and the Court cites none) that it also includes DNA from all-or even any-crimes whose perpetrators have already been convicted.
While Scalia is correct that DNA from "solved" crimes is not routinely checked against the offender database wrongfully convicted people have been exonerated after special checks. A Westchester case with which I am familiar is that of Jeffrey Deskovic. A somewhat weird kid he was convicted based on a false confession of the rape murder of a high school classmate despite that fact that he was not a match for DNA found on the victim (on the unsupported theory that the DNA was from a boyfriend). After many years in prison he convinced a new DA to run the DNA against the offender database. This hit against a prisoner serving a life term for a different murder. The prisoner confessed and Deskovic was exonerated. Now in this case the prisoner had already been convicted so under Supreme Court precedent it was okay to extract his DNA regardless of the outcome of King v. Maryland which concerns taking DNA samples from people who are accused but not yet convicted of serious crimes. Still it should be clear that people in Deskovic's unfortunate situation are more likely to be exonerated when broader criteria are used to build the offender database.