I think more it takes so long due to the amount of cases to be heard plus the appeals not so much Bongiorno's maternity leave. Rafaelle did have another attorney besides Bongiorno to be at his side.
If this length of cases helps the Italian judiciary achieve its judicial truth I will not find fault, however, many cases may not have a satisfactory verdict for either side due to the length of time it takes to hear and finally decide a case.
I believe that in a criminal court in England & Wales, if a defendant's barrister was a female who had a baby and then maternity leave (or, for that matter, someone of either gender who was justifiably indisposed - for example through a family bereavement or a serious accident or something like that), then the defendant would have the option to have the case adjourned until the defence barrister was once again able to take part in proceedings (alternatively the defendant could choose to appoint a new barrister - but I believe the choice between a new barrister and an adjournment would be entirely up to the defendant). But in either case, I don't think any court would accept (or would, for that matter, require) an adjournment of anything longer than a few months.
And, that being the case, the "argument" about Bongiorno's pregnancy/maternity being a significant factor in the eight-odd years this case took to reach a final verdict (in Italy) is somewhat misleading. As you point out, the Italian system is utterly overwhelmed, owing chiefly to the crazy system of multi-trial processes and the batting of cases to and fro between the various courts. The Knox/Sollecito case was, of course, a textbook example of the worst of this.
The Italian system could - and IMO should - easily be changed to one in which the appeal court stage is totally skipped on the way "up the chain". In other words, there's a court-of-first-instance trial, then - for all serious cases - there's no appeal-level second trial, bu rather the case moves all the way up to an automatic review of that first-instance court's verdict by the Supreme Court. And if the Supreme Court finds that the first-instance court has erred, the case is sent back to an appeal-level court for a (higher form of) retrial.
Note that in this system, the automatic SC review would only take place for the most serious of crimes: murder, rape, serious organised crime, kidnapping, etc. For all other crimes, those convicted by lower courts (or the state, if the defendants were acquitted, under certain circumstances) would have to apply for leave to appeal, and would have to present material evidence of either failures in the first-instance trial or (in the case of an appeal against conviction) new exculpatory information. So at a stroke, the workloads of both the appeal-level courts and the Supreme Court would be hugely reduced.
In addition, if there's an automatic system of appeals (as there currently is in Italy), then the State really must find a way to process those appeals expeditiously and in a just timeframe, or be forced to alter the system. Far wiser minds than I have stated - correctly - that justice delayed is justice denied. Frankly, given that there is such a huge backlog of cases waiting for appeal-level and SC-level review in Italy, the Italian state should have changed the system long ago in order to vastly reduce the waiting times between each (mandatory) trial.
(Note that the same rule applies to the European Court of Human Rights, which also has a tremendous backlog (due in no small part to the number of applications against Italy......). The ECHR should have been reformed and/or greatly expanded a long time ago. But since the ECHR is a supra-national court and since it can - and does - expedite those cases where the applicant is either currently in custody or under some other form of state punishment, the waiting times cannot be viewed as quite so troubling in most instances (including that of the Knox application).