I would use online resources to keep learning and create some side projects since it sounds like you don't have much job experience yet. Contributing to open source projects or being active on sites like Stack Overflow will help you learn too. You may think that you can't really teach with little experience, but even if you're 3-6 months in you have things to teach to beginners. Paid sites like Treehouse, Codecademy and the like will be higher quality and let you focus more than free online tutorials, but it's certainly not required.
I wouldn't focus on only what is in demand right now, take a longer view for your career. It's the thinking and planning behind programming that is really valuable, not whatever programming language is popular at the moment. Those problem solving skills will pay off in the long run more than being an expert in say, Javascript. Ask why and what you are building, not just how to do it.
This is an interesting lifestyle business, where you're essentially curating a selection of Japanese candies. However it could be hard to scale if you're just working by yourself. Great success could be your downfall if you suddenly got 1000 new subscribers at once. I would recommend thinking about what you would do if this happened, how to expand your business quickly, hiring part time workers and such. Since you're operating in Japan this could complicate things tremendously. It would be good to have a business lawyer that you could consult with.
Since you're doing the two week shipments as "random" selections, perhaps you can lock your subscription list a week or two out from the next shipment date to give yourself a set number of subscribers for that shipment. This will make it easier to acquire all of the products since your subscriber number will not fluctuate.
Then have a standard stock of easily acquired products pre-packaged and ready send to new subscribers who miss the lock date for the next shipment.
After all, I would assume that one of the big drivers of this business is novelty. For the first few shipments, it doesn't matter too much what is sent, as long as it's new and unique. This is why you will see people subscribe for a short time and then cancel (novelty wears off).
I wouldn't focus on only what is in demand right now, take a longer view for your career. It's the thinking and planning behind programming that is really valuable, not whatever programming language is popular at the moment. Those problem solving skills will pay off in the long run more than being an expert in say, Javascript. Ask why and what you are building, not just how to do it.