2. Cities can not sustain themselves but need the existence of villages that would sell them the food.
In the late medieval period farmers absolute would sell the surplus for money on the market. We also see also traveling merchants that would go from village to village to offer wares.
> Sell it to their neighbor, who is also a farmer and grows all the same things?
Some people in a village did not have land but were craftsmen like the iconic blacksmith.
An yes it was a development and early medieval period ages the serf was probably just happy to have slightly more to eat. It didn't all happen over night. Villages started very self-sufficient. Basically producing everything they needed themselves while we see much more diversification of labor later on.
3. I made a comment about the US South as well. An yes, history is not linear. There are just general tendencies.
By the late medieval period, sure, there was somewhat more robust intercity trade. But that's approaching 1000 years after the fall of the western Roman empire, which is stretching it if you want to claim cause and effect.
There was a fair bit of specialization and cash-cropping in Rome, though - it's unfair to assume that Rome and the medieval period are the same here, when the key difference is Rome, which 1) acted as basically a giant trade union, economically, and 2) Rome was built around the mediterranean, which obviously made shipping easier than in the middle of Europe.
You are picking a very specific area and comparing it to the entire Roman empire? The proportion of the population that was enslaved was significantly higher in Italy than in the rest of the empire (and it was about on the same level as in the Southern States by 1860, around 30-35%).
I can't find reliable figures but it was significantly higher in Sicily and Southern Italy and probably as high if not higher than in South Carolina.
> Sell it to their neighbor, who is also a farmer and grows all the same things?
No but there was still extensive trade on the regional level. Not by modern standards of course since surpluses were very low (but not by Roman standards since productivity had likely increased significantly throughout large swaths of the former Roman empire due to technological progress). e.g. the production of wool which was exported to the low countries made up a non insignificant fraction of the English economy.
> significant cash crop economy in medieval Europe.
I think people often tend to compare Northern and Western Europe during the middle ages with the Mediterranean centered Roman empire for some reasons. I'm not sure there was an extensive crash crop economy was particularly more extensive in Northern Gaul, Germany or Britain during the Roman period (probably the opposite since the population sizes seem to have been much higher in those regions during the middle ages).
The plantation/latifundia based economy in Southern Italy had collapsed of course but I'm not sure that the production of main cash crops wine/olives was necessarily much lower overall by the high middle ages even in Southern Europe (though of course it varied by exact period and region).
2. Relating incentives to productivity is an anachronistic grafting of industrial and post-industrial economic thought onto a subsistence farming economy where it doesn't apply. If a serf works harder and produces more crop, what are they going to do with it? Sell it to their neighbor, who is also a farmer and grows all the same things? To the best of my knowledge, there was not a significant cash crop economy in medieval Europe.
3. As others have pointed out, slavery continued long after the Roman Empire. It was much more prevalent in the encomienda system and the antebellum South than it ever was in Rome. The proportion of enslaved people in mid 1700s South Carolina about 3 to 5x that of ancient Rome. And even that's nothing compared to the Caribbean.