I brought up the time reversal invariance of the fundamental laws because I thought it would help reinforce why the 2cnd law of thermodynamics could not be a rigorous law of nature, but only a statistical one. Unfortunately, probably through poor presentation, I feel I may have only muddled the issue. It is not that the 2cnd law can only be violated if someone were to build a magical machine which could reverse time.
The time invariance of the fundamental laws shows that any process which proceeds in one direction can also proceed in the reverse direction. This is in direct contradiction to a rigorous adherence of the 2cnd law.
A direct consequence of the time invariance of the laws of quantum mechanics is the principle of detailed balance which states that the transition probability for the transition
i->j is the same as the transition probability j->i.
To quote the respected physicist Herbert Callen (Thermodynamics and an Introduction to Thermostatistics p. 468)
The equal probabilities of permissible states for a closed system in equilibrium is a consequence of time reversal symmetry of the relevant quantum mechanical laws.
This means that the time reversal symmetry of the fundamental laws leads directly to the fundamental postulate of statistical mechanics, The Postulate of Equal a Priori Probability.
The 2cnd law of thermodynamics can be derived from this postulate.
Therefore, there is no time reversal problem.
How does one then reconcile the time reversal asymmetry of the 2cnd law with the time reversal symmetry of the fundamental laws? By realizing that the 2cnd law is only statistically correct.
Since it is only statistically correct, it means that it is possible for it to be violated (and yes I mean in the forward time direction type of way). I tried to give an example of this using the kinetic theory of gases, but you didn’t like it.
When contemplating whether the universe must die a heat death at maximal entropy the physicist Kerson Haung stated (Statistical Mechanics p.19)
Our universe is governed by molecular laws, whose invariance under time reversal denies the existence of any natural phenomenon that absolutely distinguishes between the past and future. The proper answer to the question we posed is no [about the heat death of the universe]. The reason is that the second law of thermodynamics cannot be a rigorous law of nature.