Everyone is different in how good they are without any training, and in how they respond to different stimuli. Some of have a naturally high vo2max, some low. Some are able to vastly improve their vo2max over time with training, while others are low responders to vo2max training. Same thing with lactate threshold adaptation.
But, we can typically make some pretty good, educated guesses about what generally works and what generally does not work. By and large, vo2max and lactate threshold are both trainable, and we have a pretty decent idea as to how to train those things.
As a preliminary matter, I would say that it is hard to really say what has caused his improvements. He is only 24, and he has upped his mileage significantly, so it is certainly possible that a portion of his improvement is attributable to maturation and increased volume rather than steady-state running at what is probably just a bit slower than marathon pace. We also do not know how he trained before this, nor do we know if he would be a better runner if he trained at a slower pace on his non-workout days.
Assuming that some or all of his improvement is attributable to him running his non-workout days, based on the peer-reviewed science out there, I would reach one of two conclusions: either (1) running at just a bit slower than marathon pace results in an adaptation that is useful for running but has not been identified and this approach can be widely adopted, or (2) this guy is an outlier and is responding to training that works for him, but does have broad applicability to the population at large.
My guess is that his improvement has very little to do with running at this moderately hard pace (and to the extent there are benefits from this, they are mostly psychological), and to the extent that this guys is reaping any real benefit, it is likely because this guy is some kind of outlier, and just because it works for him does not make it an approach that the masses should try to emulate.