So while, theoretically, you could hunt down participants in all of its interactions and deduce meaningful evidence about its history, the technical act of doing so is practically impossible because the problem of discernment grows with the age of every other carrier within some volume; making it intractable or, at least, wasteful to the extent possible.

When we’re talking about things like following blood particles back through a river, we have fully left behind the realm of practicality and wastefulness. Information can be scrambled well beyond the point of making recovery feasible, but we’re talking about whether it theoretically exists.

One way to look at it is where the “cut-off” point would be. While the blood is streaming away from the body, we can see exactly where it’s coming from. What makes us lose track of it are the limitations of our instruments.

If you can find a trace of blood and reconstruct where it was even a second ago, then there’s no reason (apart from the practical ones) you couldn’t repeat that process and get the location a second before that, and so on.

All that said, I often think this way of speaking of entropy is somewhat unhelpful in that there are many forms of entropy and not all should obey the second law.

I don’t think there’s any way of getting around the second law, period.

Take position entropy. One way to look at it is to see how many things are in the universe at different locations; if you count all that up you have a measurement (aka volume); maybe divide by the number of things for comparability’s sake and would you look at that? It’s density. Another chap chimes in saying something to the effect of “can’t fool me, position entropy’s just ħ/2Δp summed over all event participants”. Call me pedantic, but it’s not obvious that these measurements must agree

Ngl, you lost me. Like I said, I’m rusty with this stuff.