* The article about flying inverted was most interesting-- that's a good spin on these questions. It made a pretty good case that P-factor and spiralling slipstream are both important, with P-factor tending to dominate. I assume these aircraft had inverted skip-skid balls on the panel as well as upright ones.
* A Squared, that's a good point about the flow direction being influenced by the wing, in the case of a pusher prop located behind the wing. So P-factor might be minimized or might even operate in a reverse sense. I still think the yaw effect is primarily due to spiral slipstream in these cases, with the fin being so close behind the prop. I'll have to give some more thought as to how to explore this further experimentally.
* To those who are inclined to called the spiralling slipstream "nonsense"-- a few minutes of google searching on the terms "spiral slipstream tufts visualization fin" turned up this pdf
http://www.lr.tudelft.nl/fileadmin/F.../2005_4_02.pdf entitled "Propeller Wing
Aerodynamic Interference"-- not a picture of tufts, but take a look at pages 16 (p.30 in the PDF) through 24 (38 in the PDF). The spiralling slipsteam is clearly described. Pay special attention to Fig 2.9 on page 23 (p. 37 in the PDF). The figure is showing a swirl angle of 3 degrees that appears to stay nearly constant as we move more distant from the prop.
Fig 2.25 on P. 43 (p.57 in the PDF) is kind of interesting -- how the prop slipstream changes the wing's angle-of-attack.
The experimental investigation section begins on p. 91 (p. 105 in the PDF).
See figure 5.43 on P. 138 (152 in the PDF) for a photo of deflected tufts due to spiral slipstream. This article focussed on prop-wing interference so it's not really what we want-- the tufts are on the wing-- we'd like to see a photo of tufts on the aft fuselage and tail-- but it's a start. A few more minutes of googling around would probably turn something up.