LATITUDE: The most obvious and simple influence on climate is latitude. At the 60th parallel south, you could expect freezing temperatures every day May-Sept, and some days almost every month. Here is climate data for other islands at that latitude.
ALTITUDE: "From the southern part to the center, half of the island is a 3km high plateau that end with cliff at the center of the island" – This plateau is certainly going to be a very harsh place. The effects of that altitude alone are going to lower temperatures 18-21°C from baseline, so it'll typically be -20° to -10°C. There are places where people live happy lives in that sort of cold. (Llama can also survive at such temperatures.)
EFFECT OF THE CLIFF ON PRECIPITATION: "From the southern part to the center, half of the island is a 3km high plateau that end with cliff at the center of the island" – as northerly sea winds hit the cliff and are forced upwards, they will drop their moisture, creating a wet and fertile area to the south of the cliff. This may not be the population centre (that will probably be the coast), but it may be a green plain good for llama/alpaca/vicuña pastoralism. Rainfed rivers could flow from this plain to the coast, with your population centres on the mouths of the rivers. The high plateau will be arid.
EFFECTS OF LANDMASS: Seasonal fluctuations in this central plain will be greater than at the coasts.
GLOBAL EFFECTS: The Drake passage is a key place to the global climate. This island (arguably really a continent) would block thermohaline circulation. Now you could arguably handwave pass that, especially as you said seawater flows under/through the island, and it would greatly complicates your plot, but your science is getting softer if you do.
FRAME CHALLENGE: On your first numbered point: there are no hurricanes in that area to begin with.
I also agree with others that it's impossible to answer this question in full hardness. The effects on the flow of wind and water will be enormous, and can't be computed without significant well-funded research.