Mike makes a very good case for putting emotions in front of analytics .. how when it's time to create a true breakthrough product you have to be ready to back away from data. Great products are a work of art, art is emotions and emotions defy data! After all would you as a consumer want to drive a beautiful Jaguar XK SS 1957 or a Toyota Prius?
Still, through the product pitch, what he says makes a lot of sense. White space is important and over optimization of space to get better click through, or maximum commercialisation for any product, is a recipe to lose the emotional connection with the users.
White space in Flipboard's business is content. Applied to what we do, white space in my business is to hone in on what the user needs to do and make the experience truly simple, flawless and enjoyable. Mike's principle of building simplicity and putting the task to be performed at the centre of a design is a driver of success. However, I disagree with his argument that this can be achieved 'just' through emotions. Using the car analogy is a nice trick but incorrect, because one (Jaguar) optimizes one thing - beauty; the other (Toyota) optimizes fuel consumption. It's like comparing apples to oranges and saying one is tastier than the other based on their vitamin content; he is mixing up two different business goals. I strongly believe in emotions in design myself, but great products are not just emotions. It's not data versus emotions, it's data AND emotions. Internals of 'insanely great' products in my view are the make up of data gathered through user research science AND emotions AND tamed technology challenges.
User observations is fuel to build the right emotions and usefulness into the product experience. I believe that user observations is data and it's one of the vectors of inspiration to use in my design activities to create the emotions that will connect with users and build breakthrough products.