The issue is that the answers are (yes basically) universally different regardless of whether both options are provided, but that statistically speaking they are identical.
The foible of humanity that makes the answers different is that we first "rebase" our expectations to the 100% chance, rather than considering our current position to be the baseline.
Yes, statistically they are identical. However, the fact that people consider the significance of the objects and the context of the decision isn't a fault of human intellect.
Really, stating otherwise is a fault of human analysis. It's a game theory problem I think (likely a variation of the stag hunt: http://en.wikipedia.org/wiki/Stag_hunt). It's a paradox only because Armchair mathematical intuition fails to explain it.
No because the stag hunt family of problems are about trust in other people, the situation in the OP is not. It's about risk aversion. Whether or not that is a 'fault of human intellect' is a matter of definition and up for debate, it's not as clear-cut as you make it out to be.
So you are saying that with the following scenario:
1. +3000 at P100 or +4000 at P80
2. -3000 at P100 or -4000 at P80
Depending on the context, the answer is basically universally different if you say "1 then 2" as opposed to "just 1" or "just 2"?