It is a proximate empirical observation of a phenomenon. At the quantum level, it is violated all the time. It must be violated at thermal equilibrium as well (unless at absolute 0, which would violate the third law) - so it is a statistical phenomenon seen in bound systems when regarded at a certain scale.
As an observational phenomenon, a system's degree of order or randomness is subjective. A new deck of cards is ordered. After shuffling, it becomes disordered - it is highly unlikely that the deck of cards will ever become ordered as originally after any number of shuffles. But the idea that one arrangement of the cards is peculiar is subjective, all sequences are as likely, we just think that some are unusual.
The second "law" summarizes a practical observation, and gets us to looking for an external energy source when an otherwise closed system seems to increase its order. It is useful when we want to make a reservoir for energy or extract energy from a system. (We can use clumsy methods). But have you seen a formula relating energy to the degree of order in a system? I haven't.
What about self-organizing system? Complex systems seem to form ordered structures all the time, without consuming energy. One might describe these phenomena as eddies in the entropy flow, I suppose, but it is hardly conclusive. Just because we are able to discern organization in a constrained system, I am not convinced that it means that it is energetically different from any other arrangement of its constituents.
I see a green light much sooner than the dim-bulb on the cell phone in front of me. However, sound has remained a constant in that my cars horn can speed the green light signal up such that they can see it. If the horn thing doesn't work then the physical, anti-inertial assault from my bumper to theirs will speed it as well.
Stay safe; stay armed.
PS: I think that the author is full of crap. I cannot see a mechanism that can slow light's speed except for a physical barrier. Black holes can refract light, bend light and even alter its course by 180 degrees, but not slow it down. However, I am almost always wrong!!!