See whether you think this is (1) awesome police work, or (2) an interrogation that convinced a 73-year old Mexican immigrant to falsely confess to molesting a child. I can see strong arguments on both sides.
The elderly man at issue was clearly not sophisticated (a huge understatement). The police took advantage of that by repeatedly lying to him and convincing him that because the "science" would clearly prove him guilty, his only way "out" was to say what the police were telling him he had to say. Which he then did. You can definitely see how this might result in false confessions. And it's not like there's a ton of other evidence that the guy in fact molested the four-year old girl at issue.
At the same time, maybe the guy is guilty, and the police got him to incriminate himself. Perhaps accurately.
It comes down, I think, to a value judgment about what level of risk you're willing to take that you are encouraging false confessions. (The theory that we can just admit the evidence and "let the jury sort it out" seems both a cop-out and demonstrably false.) If you care deeply about not putting an innocent person in prison, I think that interrogations like this one have to be stopped. But if you're willing to run a 10% (5%? 1%?) risk of a false confession, then I can see why you'd let this stuff go on.
Ultimately, here, the Court of Appeal holds that the present facts are pretty darn close to a prior case that held that the interrogation was custodial, so it was ineffective assistance of counsel not to try to exclude the confession (since there were no Miranda warnings). But the broader issue remains. Say the police had indeed given the warnings. Which I have no doubt would not have mattered in the slightest to what the elderly man in fact did here. Are we then totally fine with police methods like these?
