Here is a relevant point-- take an engine with clockclockwise-rotating prop (as viewed from rear) and mount it high up at the mid-fuselage like this
http://blog.aopa.org/letsgoflying/wp...New-Talon1.jpg, and what happens? Now you need to hold LEFT rudder, not right rudder, in a full-power climb at high angle-of-attack / low airspeed. Even at cruise power/ cruise airspeed, the aircraft will tend to need some left rudder, if there is no compensating trim tab.
Why? Surely because with the high engine/ prop location, the fin is primarily feeling the BOTTOM half of the spiralling slipstream, not the top half. So the aircraft tends to yaw right, not left.
Many, many ultralights and lightsport aircraft share this configuration and they all experience this effect. I have some firsthand experience with such aircraft. Take a look at the trim tabs on the rudders of such aircraft-- they are always set to cause the rudder to deflect to the left to help compensate for this effect. The OPPOSITE of what we typically see when we look at the trim tab on the rudder of a more "conventional" GA plane, with the SAME direction of prop rotation.
You CAN'T explain this with P-factor, engine torque, etc. It's got to be due to the yaw effect of the spiralling slipstream striking the vertical fin.
You can't tell me that the spiralling slipstream is not important! It certainly affects the balance of YAW torques. And through slip/ dihedral coupling, this has an influence on roll.
The balance of ROLL torques is another story. In the case of a clockwise-rotating prop, as the prop spins the air clockwise, the aircraft will tend to roll counterclockwise (left). Any ROLL torque directly created by the impact of the spiralling slipstream against the wings, fuselage, etc will take some of the "spin" out of the propwash, which will REDUCE BUT NOT ELIMINATE the net left roll torque created by the prop. It's an issue of conservation of rotational momentum. It is well explained here: section 9.5:
*9**Roll-Wise Torque Budget
PS near the end of section 8.5.2 of the same source, we read that the yaw effect of P-factor is SMALL compared to the yaw effect from the spiralling slipstream striking the vertical fin:
*8**Yaw-Wise Torque Budget
We can deduce the same by noting that even with a clockwise-rotating prop, these aircraft with high-mounted engines tend to yaw right at high power, not left. P-factor would induce left yaw, just as with any more "conventional" aircraft with the same direction of spin of the prop.
We do all understand slip-roll coupling due to dihedral, right? If the ball is significantly off-center (say to the left), that generally means the aircraft is flying a bit sideways through the air (in this case a "yaw string" would blow toward the right). That sideways airflow interacts with dihedral to make a "downwind" roll torque (toward the right in this case.) To look at any roll effects from the prop that are NOT due to yaw/ sideslip, you'll need to first start by applying whatever rudder pressure is needed to center the ball, or more precisely, to achieve zero sideslip. (Yes there is a difference between these two things, as becomes apparent when we are dealing with very large rudder deflections-- eg single-engine flight in a twin-engined aircraft-- but that's a rather fine point in the context of a single-engine aircraft. Zero sideslip is achieved with slightly less rudder deflection than would be needed to fully center the ball.)
Re the original question-- let's say we have a prop effect-- which again I contend is mainly due to the spiralling slipstream-- that makes a left yaw torque. The nose will yaw some degrees to the left of the flight path, displacing the ball to the right and then the net yaw torque will be zero. Now the aircraft will fly along at a constant yaw/slip angle-- the ball is deflected to the right. The sideways flow over the aircraft will have a SMALL tendency to drive an unbanked left turn. But the sideways flow over the aircraft will also interact with dihedral to make the aircraft roll left. As the bank angle increases, this will create a much stronger left turn than would ever be created by the yawing/slipping condition when the wings were level. Sure, if you try to fix the problem by holding right aileron, the adverse yaw from the deflected ailerons will shift the ball further to the right. You can fly in a straight line this way, but not quite wings-level, and it's certainly not efficient-- the wind/ airflow is slamming into the right side of the fuselage. If you had fixed the problem by applying right rudder to center the ball, you would never have needed to apply the right aileron input. As a glider pilot you understand the yaw string. The ball is essentially the same, it just moves the opposite way. Center the ball with the rudder before you figure out what aileron input is needed. You might be pleasantly surprised. You are interpretting the fact that you are holding lots of aileron, as an indication that the prop is mainly making the aircraft roll, not yaw/slip . That's not a valid conclusion. The prop is making the aircraft yaw/slip , and because you aren't fixing that with the rudder, dihedral is then making the aircraft roll into a banked turn, unless you hold some aileron input.
PS Sorry-- I just re-read your question-- you are stating that on these sims, holding rudder to center the ball does not leave the aircraft balanced in roll, and holding rudder to stop the aircraft from turning does not leave the ball centered. It sounds like the sims are modelling a strong left roll torque from power, even with the ball centered or nearly so. In theory there ought to be SOME left roll torque from power, that should need some left aileron to counteract (w/ ball centered), but it sure sounds like these sims may be over-doing it. But maybe that is indeed realistic for prop-driven aircraft with very powerful engines-- that's outside my area of experience.
PPS re oil streaks on the fuse etc-- it really doesn't take that much spiral in the slipstream to have a strong effect on the vertical fin. A few degrees of change of angle-of-attack of the fin will surely have a significant yaw effect.