Italian is commonly analysed as inheriting the nominative forms of nouns from Vulgar Latin, instead of the accusative ones. But what happened to 3rd declension nouns?

It looks like for the majority of them, due to the irregularity of the sg. nom. forms, the sg. nom. was simply remodelled with the sing. accusative. This explains e.g. 

(Latin noun forms in the order of sg. nom., sg. acc., pl. nom./acc.)

`voce` "voice" from `vōx, vōcem, vōcēs`, 

`-trice` "-er (f.)" from `-trīx, trīcem, trīcēs`, 

as well as `fiume` "river" from `flūmen, flūmen, flūmina`. 

Fair processes. Now, what happened to some other Latin 3rd declension patterns?

 1. It. `-ione` "-(t/s)ion" from Latin `-iō, -iōnem, iōnēs`,

    - Did the same process happen here as well?

 2. It. `-[d/g]ine`  from Latin `-[d/g]ō, -[d/g]inem, -[d/g]inēs`, 

  - incl. It. `voragine` "chasm, abyss" from Latin `vorāgō, vorāginem, vorāginēs`, It `torpedine` "stingray, torpedo" from Latin `torpēdō, torpēdine, torpēdinēs`, It. `-itudine` "-itude" from Latin `-itūdō, -itūdinem, -itūdinēs`, 

    - Same question as above. In addition, is there any particular reason why the Italian reflex is not `-[d/g]ene` given the short `-i-`?

 3. It. `-ità` "-ity" from Latin `-itās, -itātem, -itātēs`:

   - Wiktionary lists an older reflex of It. `-itade`. 
     - How did this sufix happen, given Italian did not go through the intervocalic voicing of lenis consonants?

     - Naïvely it appears there was a competition between the two forms and `-ità` won out. Is this the case?

   - It. `vertù` "virtue" from Latin `virtūs, virtūtem, virtūtēs` seems to have undergone a similar process, with an older form `vertude`. Then, how did the  `-de` drop out in this case?