Future gravitational-wave detectors, especially the Laser Interferometer Space Antenna (LISA), will be sensitive to black hole binaries formed in astrophysical environments that promote large eccentricities and spin precession. Gravitational-wave templates that include both effects have only recently begun to be developed. The Efficient Fully Precessing Eccentric (EFPE) family is one such model, covering the inspiral stage with small-eccentricity-expanded gravitational-wave amplitudes accurate for eccentricities $e < 0.3$. In this work, we extend this model to cover a larger range of eccentricities. The new EFPE_ME model is able to accurately represent the leading-order gravitational-wave amplitudes to $e \leq 0.8$. Comparing the EFPE and the EFPE_ME models in the LISA band, however, reveals that there is no significant difference when $e_0 \leq 0.5$ for binaries at 4 years before merger, as radiation reaction circularizes supermassive black hole binaries too quickly. This suggests that the EFPE model may have a larger regime of validity in eccentricity space than previously thought, making it suitable for some inspiral parameter estimation with LISA data. On the other hand, for systems with $e_0 > 0.5$, the deviations between the models are significant, particularly for binaries with total masses below $10^5\, \mathrm{M}_{\odot}$. This suggests that the EFPE_ME model will be crucial to avoid systematic bias in parameter estimation with LISA in the future, once this model has been hybridized to include the merger and ringdown.
Comment: 22 pages, 9 figures, to be submitted for publication