Howard, can can you have an accurate output when there is no accurate input?
So you are "in total agreement" with an assertion that the extra drag of a foot of lines suffices to lean out the engine, causing it to sag in a vertical eight, having come to this conclusion from "no accurate input".
Here's how I would go about evaluating some of the engineering claims above. First, do some rudimentary calculations to see if an assertion is anywhere in the ballpark. For example, take the assertion that the extra drag of an extra foot of lines causes the engine to lean way out. In this case, there are indeed some unknowns: airplane speed, the length of Greg's arm and handle, and air density. I could have asked Greg, but reasonable assumptions suffice for now. They can be refined later if needed. I assumed airspeed to be constant: that's the worst case. It would result in the greatest drag difference. If the answer you get for the difference in drag is trivial, you can dismiss the assertion and not embarrass yourself by stating it or not be led astray by accepting it. I did some calculations on the attached spreadsheet. Assumptions are in yellow. You can see the formulas in the cells. Tell me if I made any errors. Looks like drag is .0048 lb. more with 63' lines than with 62' lines. A lap time difference of 40 milliseconds would give the same drag increment. This wouldn't cause your engine to lean way out and not make it to the top of a vertical 8. Another equivalent drag increment would come from turning a Supertigre-type needle valve from fore-aft alignment to sideways if the bent part is .72 inch long (not counting the effect of changing the mixture to the engine).
The next step would be to vary the assumptions to see if they matter. We do this systematically in airplane trade studies to see which inputs we have to refine and which we don't need to mess with-- Google "tornado chart". For example, what if Greg's arm is 4 feet long? Lap time difference to get the same incremental drag would be 39 milliseconds, rather than 40. You can fiddle with the inputs on the spreadsheet to see what happens, but I think the extra-drag explanation can be dismissed.
When a model flys on longer lines to achieves a higher true airspeed due to flying closer to straight line.
From the spreadsheet you can see that the incremental angle from the different circle curvature at the prop disk is .01 degree, equivalent to an incremental crosswind of .017 ft./sec. The incremental angle from the different circle curvature at the rudder is .02 degree, equivalent to an incremental crosswind of .028 ft./sec. I don't know about your engine, but I don't think .028 ft./sec. of wind would make any of mine go nuts. Maybe you think that the curvature has some kind of dissipative effect, such as that which will cause the earth to crash into the sun in a couple years.
The bigger loops will result in drastically reduced G force during the maneuver and noticeable change in engine performance.
Drastically? Come on, Paul, you're an engineer. You'd get an equivalent G force reduction from going 39 milliseconds slower per lap. The engine wouldn't notice.
Igor's rudder compensation would require an incremental .05 oz of side force on the vertical tail. Don't overdo the rudder tweak.