Earlier quoted context omitted.
Macroscopic states are not a complete description of physical systems and the same microstate could be part of different macroscopic states. The thermodynamic variables are not properties of the microstate (the "physical reality"). They are properties of the macrostate, which is a model of the unknown "physical reality". If you don't like the original phrasing, let's say that "Thermodynamical entropy is not a fundame…
How could the same microstate be part of different macrostate? Like, if I described the microstate of a drop of water, wouldn't it have, as a whole, one temperature? I thought that a macrostate could have many different microstates, but not vice versa
The energy of the microstates corresponding to the macrostate is not fixed. The probability of each microstate is proportional to the exp(-E/kT). https://en.wikipedia.org/wiki/Boltzmann_distribution
The microstates don't have a temperature: they are more or less likely to happen depending on the temperature.
And that's assuming a particular thermodynamical model. The same microstate could be looked at in the context of a different thermodynamical model as a different (or differents) macrostate(s) characterized by another set of thermodynamical variables.