I think I already know the answer (Pablo may) but I put in a email to technical. I'm behind on my reading.
What it should say (I believe) is: ATL recommended wherever WS is specified EXCEPT when it is specified for a CVT transmission (one statement).
Putting "Not for use in CVT transmission" as a side note while recommending it for (otherwise) all WS applications just ..well, it's ...it's just rude.
The Prius does not use a CVT in the traditional sense. The gear ratio of the PSD (Power Split Device) of the Prius is fixed and cannot be changed. Instead, the Prius uses it's two electric motor/generators to change the power loading and flow. In many cases the some portion of power from the engine is routed *electrically* around the gears via the electric motor/generators.
There are no clutches or frictional surfaces in the Prius transaxle. There are many cases of many different people over on http://www.priuschat.com using Redline D6, or Dexron VI, and several other fluids with UOA showning good results.
The only negative was a high copper wear metal with Amsoil ATF in the first gen Prius transaxle, which from the factory specs Toyota IV in contrast to WS being spec'd for the 2nd gen Prius. I do not know what is spec'd for the latest (2010+) 3rd gen Prius.
I would say use it with confidence.
We are talking about ATL, not ATF for a WS application, not T-IV.
I think the test above is pretty cool, but inconclusive. It's well known that ester bearing synthetic fluids will hold Cu in solution, plus he had (and reports it as such) Si/abrasive action, not related to the ATF.