Of course the first thing I had to do was visit the author's website to find out what exactly they are selling. What I see here, is what I have seen before. Building a neat tool for educational purposes, or leaving it up to the customers to determine the commercial applicability. Once they come up with something marketable, you do well by being their supplier.
Relying on your customers to come up with the breakthrough may seem like a smart idea, but my guess is that it's a crowded marketplace. If I had venture capital to burn, I'd rather invest in a company that was making a product that solved some actual problem, with people willing to buy that product that will solve their problem directly.
Doug, you're right, this could be done anywhere! I meant for the focus of the article to be more on the potential for new development tools, not the location for it. I know companies who do this sort of thing in the Midwest (my home region) because the cost of living is lower. I know companies that have moved across the country to be in Silicon Valley because their customers are here. Either approach can work.
Yes it would be just wonderful that robots become available to every child or person. This is a wish list. Let us see when it becomes a reality. With changing times the way embedded development happens is changing, it is going to be more flexible for people wit differnt background to start embedded development.