I agree with your analysis here.
And to me, Mignini's remarks strongly support the belief that the criminal justice system in Italy is still horribly tainted by its messy and clumsy amalgamation of inquisitorial and adversarial systems.
To me, Mignini's remarks hark back to the inquisitorial days when the PM found the "judicial truth" in any given case: the PM was supposed to be a "neutral" investigating authority, with no vested interest in any particular position. Therefore, the version of events that the PM placed before the court was meant to be the "objective" assessment of the crime. And, as Mignini alludes to, the only seeming requirements was that the PM's version of events a) fit with the known evidence, and b) was not self-contradictory in any way. If it passed those threshold tests, then it was the version that tended to be accepted by the courts.
Of course, it doesn't take a genius in jurisprudence to see how horribly flawed this sort of methodology is in the application of justice. It's why the inquisitorial system is totally unfit for purpose in an age where investigative tools are so sophisticated and varied, and where human rights have actually evolved to something close to acceptable levels. But for clarity, here again are the two main reasons why this sort of inquisitorial approach is so fatally flawed:
1) PMs under an inquisitorial system still quite clearly have a balance of interest in obtaining convictions. It's therefore arrant nonsense to suggest that their presentation of the case is objective and disinterested.
2) There is zero benefit of the doubt given in practice to the accused under an inquisitorial system. This is patently unfair and unjust when the accused a) is up against the power of the state, b) is the one who stands to be punished if convicted, c) is the one against whom the accusations are made (and it should always be incumbent upon the accuser to prove the accusations), and d) can often not prove his/her innocence (which should never be an impediment in and of itself to being found not guilty).
Mignini's remarks only strengthen my belief that one of the fundamental structural and institutionalised problems in Italian criminal justice is this failure to abandon many of the fondly-held tenets of the old inquisitorial system. And the problem is that this can actually result in a far worse system of justice than even the old pure-inquisitorial system. I think that courtroom judges and PMs are still far, far too close to each other, and that court judges are still clearly inclined to place the PM's version of events as the default "starting position" - thereby shifting the burden onto the defence of essentially having to show the court how and why the PM's version is incorrect. And if the PM's version is logically consistent and is consistent with the evidence (even if, as you say, it's far from the best fit with the evidence), then it effectively becomes virtually impossible for the defence to sway the court.
I think that this is, to all intents and purposes, exactly what has happened in the Knox/Sollecito case. I think that the "burden of disproof" was implicitly placed upon the defence. I think we see it in the SC ruling and in the Nencini judgement, and we most definitely see it in Mignini's remarks. And it's massively destructive to the fair and just application of criminal justice in a modern liberalised democracy.