CAUTION: delta-wake/delta-Cd 'linearity'
 

I've been looking into Jeff Howell et al.'s wake vs Cd 'linearity' claim.
Shooting from the hip, it appears that in order for this to be 'true' would require that their 'pre-boat-tailed' test vehicle had already achieved the flow deceleration-rate saturation point ,aft-body slope, and the actual tail surface was a simple, conical section of that slope, as a constant over the remainder of it's length.
W.A. Mair' boat-tail achieves saturation @ 44% of total tail length, and out to 100% of length, the drag falls arithmetically with reduction in wake area, along it's fixed, 22-degree slope.

Wing / Strut Section Wake vs Cd Behavior
 

In Fachsenfeld's 1951 book, he shares a research he and Kamm conducted at the FKFS, very similar to what Hoerner published in his Figure 3.26, regarding the 2-D flow, Gottingen 460, 20.5% thickness, airfoil section.
The FKFS model is a 2-D flow, 8-piece, composite, K-form, symmetrical airfoil, which represents measured Cds, ranging from, zero-aft-body, to 100% aft-body, representing 8-different 'wake areas', along -equal-length segments.
The Cd varies from, 0.58, to 0.065.
--------------------------------------------------------------------------------------
What's germane to Jeff Howell et al.'s research is, the first 75% of the aft-body, which demonstrates a perfectly linear drag relation along its slope, as a function of wake area reduction, along the 'wing's' curved contour.
The last 25% becomes non-linear, with diminishing returns, as it goes asymptotic near 100% length and zero-wake.
--------------------------------------------------------------------------------------
I've misplaced Kosin & Lehmann's streamlined body of revolution research. I'll try and hunt that down, as I'd like an example of a purely 'compound-curve' - surfaced tail.