Buying decisions can have so many confounding variables (emotion, status, cost, quality) this article draws some very loose conclusions on relatively little data. Less choice can be an interesting A/B test for a business, but I wouldn't be surprised if results feel more random than this article suggests.