If the airframe is in a sideslip, the flow field has a cross-fuselage component. If the aircraft is symmetrical about it's vertical cross-section this might not have an effect, but any asymmetry such as the wing being above or below the fuselage centerline or the bottom of the fuselage being square as opposed to the top being round, etc., will manifest itself in a rolling moment. You have to visualize the cross-fuselage flow. The airflow splits and goes both over and under the fuselage. This is where the roll comes from.
If it's actually slide-slipping, the effect is at the tips, not the fuselage, and not from interference from the fuselage to any great degree. In general, however, flying in a circle doesn't cause a side-slip, in fact, absent intentional yaw, the fuselage is tangent to the circle, which negates either of these effects. You aren't "dragging the model sideways" relative to the air, which is why Dennis' statement it wrong.
Yaw (with respect to the air) certainly does change the lift and cause roll motion, in fact it is critical to trimming (and also alluded to in my suggestions, although not explicitly). There is substantial dihedral effect from a variety of sources, although I think the "blanking" effect is negligible. Flow across the tips is also altered, and that has a huge effect, both from changing the lift distribution differently inboard and outboard, and by virtue of having a large arm over which to work. That's what can roll the airplane 20 degrees and back again in the space of 1/4 second in a corner, for instance. Estimate what sort of torque it takes to roll the airplane, say, 20 degrees in 1/4 second. Then see how much force has to be applied at a point, being optimistic, 3" from the CG. I think you will find it takes far more lift/force than you use to do the entire corner. Then, do the same thing applying the torque to the tips at maybe 28" from the CG. The effect of turbulence coming off the fuselage (which is what really happens - not "blanking") is like putting a tiny tab at the root of the flap.
It is quite critical to control the yaw angle, either passively (like having 3 billboards in formation) or actively (Rabe rudder) to remove this effect. Yawing itself, if you ignored the roll effects, probably doesn't matter that much, but it sure does matter when you consider the coupling in to roll. Igor Burger, among other notables, trims to fly with substantial intentional yaw, and all these effect have to be considered. In fact, in the case of substantial static yaw angle, you also have to consider dynamic and kinematic effects associated with the inertia tensor. I have seen absolutely no one successfully deal with this consistently, although I have seen various people try it and get good results for specific conditions. When the conditions change, they usually start over or have massive problems.
While we are at it, the other assertions from Dennis' troll/cyberstalker post- Almost all competitive airplanes from the last 30 years have unequal span wings with longer inboard panels. Not the excessive asymmetry of something like an All-American, but still anywhere from 1/2 to 1" or so is almost ubiquitous. The reason is obvious, if you don't that, you end up with massive amounts of tip weight to correct for, dare I say it, the extra lift on the outboard wing. A typical equal-span wing airplane will take around 2.5 - 3 ounces of tip weight, and adding 3/4" of asymmetry will change that to something around 3/4 ounce. 3/4" is about what is required to balance the lift from side to side (i.e. put the lateral CP in the center) and 3/4 ounce is about the contribution of the lines to the lateral CG, with some assumptions.
Many times these designs also include wider flaps, which are there to balance the lift- at varying load factors, not due to speed differential, etc. If Motorman's airplane flies level at 1 g (level flight), rolls out slightly at 3 gs (round loops) and even more at 15 gs (corner), it probably does need more lift on the outboard- but only at higher load factors. That's almost certainly because we are adding asymmetry to the wing to compensate for the speed differential, but not the tail. Look at which way the tail lift rolls the airplane if the CP is outboard of the CG - exactly the same direction as would be compensated for by the tab. The tab does almost nothing level flight, not too much in round loops, but a lot in the corners.
Note that this is NOT true with equal-span wings in most cases. The one thing they have going for them is that they do not compound the tail CP issue. And they are far less prone to requiring lots of outboard flap tab - in fact, a very experienced modeler recently required a tab on the INBOARD flap because he built in too much outboard flap differential with his equal-span wing. Of course, this is like shifting the tail too far, since the lateral CG ends up in the right place WRT the wing but INBOARD of the CP of the tail, which is like there is a reverse speed gradient. The tail should have about 1/4-3/8" of differential, and what is actually happening for equal-span wings is that the tail is being shifted too far instead of not enough.
Of course, experience has shown that you can trim either equal-span or asymmetrical/correctly designed airplanes about as well. But to do it, you have to know why it is doing what it is doing, and not some strange "stunt lore" about it. I know people love their lore, sometimes over clearly documented facts, and Dennis has taken it upon himself to interrupt or derail many successful threads (note that Motorman's plane is flying well now - which was the original topic..) to try to troll me and play his games, but it really does matter what is happening, and people will never actually understand trimming unless they have a good understanding of that is really going on.
Brett