Right, it makes no sense to me to convert on board mechanical energy from AWE to electrical energy when you want mechanical energy to drive your boat at the end. So let’s add a flywheel to the boat, and batteries for eventualities.
If you’re going to be relying on wind to power your boat I don’t think you can go in a straight line, like you couldn’t in the age of sailing or can now in spaceflight.
Let’s also start from scratch: you want to buy a boat, add and delete some stuff, and then use that to deliver goods or people to another location at a speed and cost that people are willing to pay for, eventually. Your advantage is that you don’t have fuel costs. You rely on software to plot the best course, just like you rely on software to fly your kites.
You can charge your batteries at way-points, you can charge your flywheel at way-points and during sailing.
Let’s assume you’ve built the prototype and it is working. Now you’d like to be able to plot the most efficient routes based on the weather at that moment, the quickest routes, determine the locations of way-points, the size of your batteries and flywheel, your cargo. Ideas for that might come from shipping, sailing, aviation, spaceflight, and weather prediction, for example.