OK, here we go (since I already have dived DEEP into this in the past ).
First of all, never EVER stick something inside the fan/belt area with engine running, I don't care what mates have said in the past, you are an idiot if you do that, and from work experience I can see a lot of dumb idiots coming into my ED. There is a big sticker for those who have problems reading that shows a hand, a moving fan and a no-not-do-that interpretation. So far the "this needs to be said", now move on
There are several different opinions on how to test the viscous when. Especially when it should be moving with less or more resistance and how it should behave when cold or warm.
The fan clutch operates on a principle that it will lock when warming up, but spins more freely when cold. There are some that say, it should spin more freely when cold and not when hot....and there are others that say the opposite. I went through a lot of info online as well as the FSM to sort this out.
From the FSM: "If the fan-assembly free-wheels without drag (the fan blades will revolve more than 5 turns when spun by hand) replace the fan drive. This spin test should be performed when the engine is cool"
However, that, in my opinion is extreme malfunction, spinning that freely. But there is no clear way of telling suboptimal fan clutch performance, because it usually does not fail in a black-or-white manner, it just looses its 'power' over time and what you might notice is that temps creep up when the cooling is dependant on the airflow created by the fan, at idle and in slow moving traffic that is.
I, for long, kept telling myself my fan clutch was fine, and it was because it never overheated in slow moving traffic or at idle (it likely would if the clutch was bad)....till I finally replaced it during a coolant overhaul. The new clutch brought temps eventually back to where they always had been when I got it 5 years ago.
They are relatively cheap, so when in doubt or when its been many years, replace it. Simply because testing and effectiveness are not correlated in its case. I've done the research on it, hoping to find a 'bottom line answer'. But the above, well, that just my input.