To the OP, here goes the thinking behind the plan (as far as I can tell, others might quibble with some of the details).
They (and the people like Marius Bakken that pioneered this style of training) were thinking critically about how to build the most race-relevant fitness over long-term.
They were searching for a system that would be sustainable (so they could keep piling up fitness month after month, year after year) while still keeping them pretty close to race-specific fitness (so they could, again, spend a lot of the year accumulating pretty durable fitness - race specific fitness tends to be less durable).
They encountered this basic idea (it is not universally perfect, especially at very fast paces, but you'll get the idea) - on a per-mile basis, the faster the pace, the more stimulus/adaptation, BUT the stimulus/adaptation is also a product of the volume of work/duration spent at that pace, and the faster the pace, the less volume/duration the athlete can handle. They were searching for the combination of pace(s) and volume/duration that athletes could handle while maximizing the long-term fitness benefits. They discovered that they could effectively run two sessions per day in the neighborhood of lactate threshold pace, and do so a couple of times per week (lots of sample workouts have already been referenced). These paces were fast enough to produce good stimulus/adaptation, yet slow enough so they could handle A LOT of it, thereby producing "the best" results.
Like any good training program that has a focus, they realized that to support the core of their work (the LT sessions), they needed some work below (faster, higher intensity, higher lactate, more neuro-muscular demands - hill sessions!) and above (weekly long-but-not-too-long run of sustained easy effort) that pace, so they incorporated both into the weekly schedule.