The Machine Malfunction results cited below, taken from the link above provided by Diocletus, are great. I say that because it can be a basis of the product liability lawsuit that I advocate Knox bring in a US court against the manufacturer of the DNA testing machine, the manufacturer of the test kit, Italian scientific police lab, and Stefanoni by name.
If Knox files suit, the Italian scientific police organization and Stefanoni will have to hire a US lawyer to appear in the US court to claim that the US court lacks jurisdiction over them and to request they be separated from the suit. US, Italian, and British journalists will line up 3-deep to cover the story - especially if Knox's suit asks $100 million in damages.
The manufacturers will not escape from jurisdiction, assuming they are US entities or foreign organizations with a nexus/ business presence in the US.
The machine manufacturer and test kit manufacturer will have to defend themselves from the product liability lawsuit. Most probably by testifying/proving in court that the Italian police lab misused their machine or test kit and that the results obtained from it are scientifically false, and perhaps the result of lab fraud.
Which then raises the question of what did the manufacturers do about it when it became public that the results obtained were possibly or probably false. Did the manufacturers notify the court and all parties involved that the machine has been misused? I see a product liability judgement in Knox's favor. Did the manufacturers revoke any licenses issued to the lab? If not, why not? Answer: didn't know, or did know but chose to do nothing in order to protect potential future sales in Italy.
The following is from the link:
3. Machine Malfunction
Not only is contamination in the Real Time run for Batch 5, but it is also evident that the machine was malfunctioning, likely due to a failure to maintain it properly. The dilution standards (i.e., dilutions used to calibrate the machine readings) for the quantification run are shown at wells A1 through B9, and when the Ct results for these standards are compared to other runs, it can be seen that the dilution standards for RT-qPCR Run 570 are approximately 4 Ct higher than for most of the other runs (there are a few prior RT-qPCR Runs for which the machine also appears to be malfunctioning). This discrepancy demonstrates that there is a problem with the operation of the machine. Notably, Batch 5 was the last batch for which the lab used the ABI 7700 Real Time machine: all subsequent samples were quantified via a Fast Time 7500 machine, demonstrating that the lab was aware of, and reacted to, the malfunctioning real Time machine.