The magnification effect due to gravitational lensing enhances the chances of detecting moderate-redshift ($z \sim 1$) sources in very-high-energy (VHE; $E > 100$ GeV) $\gamma$-rays by ground-based Atmospheric Cherenkov Telescope facilities. It has been shown in previous work that this prospect is not hampered by potential $\gamma-\gamma$ absorption effects by the intervening (lensing) galaxy, nor by any individual star within the intervening galaxy. In this paper, we expand this study to simulate the light-bending effect of a realistic ensemble of stars. We first demonstrate that, for realistic parameters of the galaxy's star field, it is extremely unlikely (probability $\lesssim 10^{-6}$) that the direct line of sight between the $\gamma$-ray source and the observer passes by any star in the field close enough to be subject to significant $\gamma\gamma$ absorption. Our simulations then focus on the rare cases where $\gamma\gamma$ absorption by (at least) one individual star might be non-negligible. We show that gravitational light bending will have the effect of avoiding the $\gamma-\gamma$ absorption spheres around massive stars in the intervening galaxy. This confirms previous results by Barnacka et al. and re-inforces arguments in favour of VHE $\gamma$-ray observations of lensed moderate-redshift blazars to extend the redshift range of objects detected in VHE $\gamma$-rays, and to probe the location of the $\gamma$-ray emission region in those blazars.