This article is orbiting around a lot concepts that have existed in academia for a long time. There, it's called the 10-foot user interface because you use it from 10 feet away.
I like the explicit discussion of technical limitations — overscan, color reproduction, etc. But I wish there was more discussion about why you make certain decisions about font size, for example, or thumbnail size. While a grid is certainly a useful design pattern, why stick with that? Is there something that might work better for your application?
One of the big issue with a 10-foot interface is striking a balance between showing the right information without showing too much information. This requires a lot of understanding about what your specific users rely on to make their decisions, and perhaps some more nefarious research into how you can shape your users' decisions based on what information you show them.
I happened to write my master's thesis on incorporating social information into 10-foot interfaces and quickly discovered that a) people rely on different information to make their decisions and b) the relative size and space limitations mean that you can reach information overload.
The author discusses clearly (very clearly, the text is bolded in each instance) why certain design decisions are made. Typography guidelines for size and weight, color/visual embellishment hazards, and the grid's structure were determined through repeated testing and the limitations & standards of the hardware itself.