> In other words, it seems as if you can take any past state and fold it into one large concatenated present state - and, by that logic, call basically anything

> In other words, it seems as if you can take any past state and fold it into one large concatenated present state - and, by that logic, call basically anything a Markov chain? Essentially yes. Given complex enough state, more or less any process is Markovian. Some do require infinite state, which would be maybe stretching the definition a bit. In practice Markov formulation may not be a very good analysis perspective if the required state would be very large/complex.