From the article:
But ultimately, being a very expensive product without enough functionality to back that up is what killed Brilliant.
I agree: Brilliant wasn’t a great product that failed because the market was too small. It was a great idea with a mediocre execution at a wildly high price point for its target market.
One example noted in the article: it could be hard to use in the dark. Which for a light switch does present some problems. 
Also, the company executive (by his own words, an “overly optimistic” person) misstates the situation customers will face if the company shuts down. He says local integrations will continue to work as long as the other company’s API is available. But they will only continue to work as long as the other company’s API is available and remains backwards-compatible. The first required update to the integration will break it. 
will release the next generation of its smart home control panel later this year.