Due to its frequency-dependent penetration depth, eddy current measurements are capable of mapping near-surface residual stress profiles based on the so-called piezoresistivity effect, i.e., the stress-dependence of electric conductivity. To capture the peak compressive residual stress in moderately shot-peened (Almen 4-8A) nickel-base superalloys, the eddy current inspection frequency has to go as high as 50-80 MHz. Recently, we have reported the development of a new high-frequency eddy current conductivity measuring system that offers an extended inspection frequency range up to 80 MHz. Unfortunately, spurious self- and stray-capacitance effects render the complex coil impedance variation with lift-off more nonlinear as the frequency increases, which makes it difficult to achieve accurate apparent eddy current conductivity (AECC) measurements with the standard four-point linear interpolation method beyond 25 MHz. In this paper, we will demonstrate that reducing the coil size reduces its sensitivity to capacitive lift-off variations, which is just the opposite of the better known inductive lift-off effect. Although reducing the coil size also reduces its absolute electric impedance and relative sensitivity to conductivity variations, a smaller coil still yields better overall performance for residual stress assessment. In addition, we will demonstrate the benefits of a semi-quadratic interpolation scheme that, together with the reduced lift-off sensitivity of the smaller probe coil, minimizes and in some cases completely eliminates the sensitivity of AECC measurements to lift-off uncertainties. These modifications allow us to do much more robust measurements up to as high as 80-100 MHz with the required high relative accuracy of +/-0.1%.