That PCA is from the Greek study, and it's the opposite, it's some Sicilian who goes further north. In that PCA Sicilians range from Peloponnese (in that specific PCA Peloponnesians are south of TSI/central Italians) to south west of TSI average slightly towards Sardinians. Sicily was repopulated in the middle ages after the Norman conquest, even with colonists who were from northern Italy.
TSI in that study overlaps more with Italians, who are north Italians from Lombardy and Venetians. There is even some Venetian who plots in the TSI cluster. Well, I've never seen a Venetian on gedmatch who gets as first population "Tuscan", they are usually even more north-eastern than Lombards.
Once again, ask yourself what a specific PCA represents, not all PCAs are based on the same. Unfortunately the Greek study is not very clear. It seems too interested in showing a similarity between Greeks and Italians to distance themselves from other Balkan peoples.
A Tuscan relation to the Ancient Greeks? Very little, except a very very few Tuscans who could descend from the Mycenaeans/ancient Greeks and a shared neolithic/bronze age mediterranean substratum. From an autosomal point of view when two samples show similar amounts of ancestral components does not imply that these components have exactly the same origin.
