Would humans on Reboxetine do better on a forced swim test? Because if they do, but they are still depressed then that's clear evidence that the forced swim test is worthless.
On the other hand if it has no effect, then humans are different from animals and don't react the same way to drugs, but the test may still be good.
Worthless for measuring antidepressants, maybe. But I'm sure you could find sponsors for a drug that causes someone to irrationally keep fighting, even if it has no effect on mood.
The real problem here is that the field of psychiatry is bandying about diagnoses like "depression" as if they actually mean something, when actually they reflect behavioral traits which could arise from a variety of etiologies.
You'll have to go back to the earlier half of the 20th century to see research like that.
So you could make a rigged test. For example:
Tell people you will pay them $200 if they can find the exit of a maze (either real or on the computer). But the maze is rigged, and has no exit. Then time how long it takes before they give up.