Example error in law #1:
The infamous "osmosis" argument, and its stunning misunderstanding/misapplication by Nencini's court.
In general jurisprudence (and explicitly enshrined in Italian law), the rules of evidence state that each individual piece of evidence must be tested for reliability, credibility and accuracy before it is allowed to form part of the case. If it passes this test, then it can be considered alongside other pieces of evidence in forming an overall assessment of guilt/non-guilt.
That's what "osmosis" means. "Osmosis" doesn't magically strengthen the reliability/credibility/accuracy of any individual piece of evidence. Rather, it means that even if each individual piece of evidence (assuming of course that it has passed the reliability test) doesn't necessarily point in a certain direction, all the reliable pieces of evidence, taken together, might indeed point inexorably towards guilt (or non-guilt).
Example: Mrs A is found shot dead with a .38 gun outside her office building. Mr A is a suspect. There are three different pieces of evidence: 1) Mr A was heard by neighbours having a violent argument with Mrs A the night before the murder; 2) Mr A owned a .38 revolver prior to the murder, and that gun is now missing; 3) Mr A cannot account for his whereabouts in the hour surrounding the murder.
Now, each of these items of evidence, considered in total isolation, would almost certainly not constitute sufficient proof that Mr A was the murderer. But all three, considered together, clearly make it far more likely that Mr A was the murderer. THAT's "osmosis" in action.
But that's true ONLY if each of those three items of evidence is itself proven to be reliable and accurate. Suppose instead that the alleged argument might in fact have been an entirely different couple in the same apartment block. Or suppose that it's unclear whether Mr A really did own a .38 at the time of the murder: he claims to have got rid of it several years ago and none of his gun club friends ever recall seeing it in the previous five years.
If each piece of evidence is itself unreliable, the "osmosis" effect will not suddenly miraculously make it more reliable! The court in the above example would not be entitled to say something along the lines of: "well, we don't really know whether or not it was Mr A and Mrs A having the violent argument, but when we consider it alongside the other evidence, we are entitled to reason that it probably WAS between Mr and Mrs A".
And that appears to be exactly the sort of thing that Nencini's court is doing in its reasoning. It ought to be immediately appealable.