If I remember correctly, and I need to check, I think the back-of-the-napkin estimate has always been the drag coefficinent changes by .0023 X (ratio of the flap chord to the wing total chord) X flap deflection in degrees.
What I'm not sure of is of it's .0023 or .00023. Makes a bit of a difference!
Since there is a direct relationship between the lift and drag coefficients (related to 1/Aspect Ratio) it should be pretty easy to estimate using 2*Pi for Lift Coefficinet vs angle-of-attack and corrected for a finite wing, again a function of the Aspect Ratio. Square that product and divide by Pi*Aspect Ratio*.8 and you should be pretty close on the drag coefficent for the wing before the flap deflects.
Again, I'm trying to eat a yogurt with one hand, type with the other and remember all at once, so let me check my formulas before you pick up the Xacto.
But again notice that I really don't care about the airfoil - it's the ASPECT RATIO that's driving the bus.
We operate models at such rediculously low Reynolds numbers that there generally isn't sufficient energy in the flow to keep it attached, so it's pretty safe to say a separation bubble develops and most wind tunnel data would be irrelevant.
Chuck