SSDD's idea seems most like the concept of the caloric. Prior to 1900, before Einstein firmly established that atoms do exist, the concept that materials contain some sort of "fluid" which accounts for "heat energy" was still a valid consideration. Both concepts, the caloric idea and the statistical mechanical idea were held to some level of believe by the scientific community. As the existance of the atom wasn't firmly proven by experiement, the concept that heat was the result of the average kinetic energy of individual molecules was not firmly established. Lacking definitive proof of molecules, the concept of the caloric still had some standing. It was Einsteins paper of brownian motion that proved, beyond any doubt, that molecules and atoms exist. This buried the concept of the caloric permanently and statistical mechanics became the understanding of thermodynamics and heat transfer.
At a molecular level, the energy we understand as "heat" is, in fact, the average kinetic energy of individual "ultimate particles", molecules. This energy manifests itself most apparently as expansion in volume of materials. If we put a volume of mercury in a tube, as temperature increases, the volume increases. As the cross section of the tube is fixed, it is the height of the mercury that increases. And, given the well known formula of PV=nRT, at standard temperature and pressure, V=nRT/P. V = height time cross section area. Height = nRT/(PA). Bingo, we have a thermometer.
The reason this works is because the kinetic energy of the molecules works against the opposing forces of a) gravity and b) external pressure.
Classical thermodynamics attempts to describe the general behavior of large volumes of ultimate particles. Heat is the average kinetic energy of the molecules as they go wizzing about, banging into each other and the walls of the container. Entropy quantifies the number of configurations that the energy can take on over sufficient amount of time to be statistically relavent.
All of the molecules in a volume of gas can be arranged in a finite number of three dimensional configurations. The energy can be spread between the individual molecules in a limited number of ways, with some being a bit slow, the majority at the average speed, and some being very fast.
At any moment in time, it is possible that all the molecules will be bunched up into one corner of a volume. It is possible that they will all be bunched up in any of the other corners. As far as location in the volume is concerned, the only restrictions are a) no two can occupy the same space at the same time and b) Heisenberg's uncertainty principle as applied to the smallest increment of volume, to which there is only so much in a larger volume.
At any moment in time, any particular molecule can have a kinetic energy from zero to infinite, constrained only by the fact that, a) on average, the total energy of the system is what it is and b) Heisenberg's uncertainty principle as applied to the smallest increment of kinetic energy, to which there are only so many increments over the total spread from zero to infinity.
The result of all this counting of arrangements of locations and velocities is entropy, a count of the total number of ways that the ultimate particles can be arranged. S=k*ln(#ways).
That said, at the boundary of a volume of gas, the molecules are banging away at the boundary, which we assume to be an infinitely thin membrane of no affect, which is then banging away at the external molecules. When a fast moving molecule in the gas collides with a slow moving molecule in the external environment, the fast one gets slower and the slow one gets faster, in accordance with the dynamics of collisions where momentum and energy are conserved. And, energy is transfered across the boundary, moving out of the system. When a slow moving molecule in the gas collides with a fast moving molecule in the outside environment, the energy is transfered the other direction.
On average, there are more fast moving molecules in a hotter body than in a colder body so the net transfer is from hot to cold. At any instance, there are numerous opportunities for energy to be transfered from the cold to hot body, just not as many as the other way round.
As the temperature increases or decreases, the amount of energy increases or decreases, thus the number of ways it can be arranged increases and decreases. Thusly, the entropy increases and decreases. Entropy is, though, not a thing. It is simply a statistical accounting of the number of possible arrangements. At no instance of time are all of the number of possible arrangements expressed.
As such, there is no caloric, no "heat fluid". There is no "heat field" and no instantaneous property of matter and energy that moderates the direction of transfer of instantaneous energy that is dependent upon the overall energy of the system.
Entropy is simply a statistical accounting and helps describe the statistical and probabilistic tendencies of large volumes of matter when it comes to the movement of energy in and out of the system.
It simply points out that, over time, energy gets spread out equally amoung all the possible ways that it can get spread out.