but I think it unfortunate that you have used 'evolution' here for an easily reversed social change over a few hundred years whereas for all the rest of the thread we have used 'evolution' to refer to essentially irreversable biological changes of thousands of years.

There are a number of pitfalls (or pratfalls) to avoid in this kind of usage. First, be clear that evolution is not always (or even not often) "progress" from a human point of view. Deadly infections evolved from something less deadly to humans- perhaps because they "found" a way to spread more readily, like making humans sneeze. The fact that the toxins that cause the sneeze eventually kill the host is invisible to the virus.

Second, be clear about what is evolving - in the case of ideas, it is the ideas themselves, and not necessarily the society in which they thrive. The fitness to survive (or "infectiousness") of an idea can be completely at odds with the wellbeing of the members of the society hosting that idea. Just read back down this thread to see a nasty infection taking hold of the USA, or over to France for a similarly nasty one in a different social group.

Thirdly, don't fall into the trap of thinking of evolution as making entities that were the starting point for a development obsolete. The bacteria are ever with us. Never get fooled into thinking of evolution as a linear progress from primitive to modern- that cliche parade from crawling monkey, through shambling ape and stooped caveman, to upright, white, bearded, male modernity (all with the right leg forward to hide the naughty bits).

And as you correctly pointed out, the probability of reverse evolution back to the starting point is near enough zero to make no difference. The reason being that the probability of evolution of any given organism is similarly really zero- the combinations are so myriad, and the environmental contingiencies (that include every other organism on the planet among a host of other factors) so unpredictable, that evolution can only ever be viewed in detail backwards. Of course, it's not a paradox at all- you can easily set up an equally improbable situation by tossing a coin a thousand times and recording the results. The probability of that sequence is one in two to the power of a thousand - some estimates of the number of atoms in the Universe are less than two to the one hundred. But the probability of SOME sequence was almost exactly one. And the probability of the next thousand tosses repeating that sequence in reverse is almost exactly zero.