Context. The design of modern instruments does not only imply thorough studies of instrumental effects but also a good understanding of the scientific analysis planned for the data. Aims: We investigate the reliability of Milne-Eddington (ME) inversions of high-resolution magnetograph measurements such as those to be obtained with the Imaging Magnetograph eXperiment (IMaX) aboard the Sunrise balloon. We also provide arguments to choose either Fe I 525.02 or 525.06 nm as the most suitable line for IMaX. Methods: We reproduce an IMaX observation using magnetoconvection simulations of the quiet Sun and synthesizing the four Stokes profiles emerging from them. The profiles are degraded by spatial and spectral resolution, noise, and limited wavelength sampling, just as real IMaX measurements. We invert these data and estimate the uncertainties in the retrieved physical parameters caused by the ME approximation and the spectral sampling. Results: It is possible to infer the magnetic field strength, inclination, azimuth, and line-of-sight velocity from standard IMaX measurements (4 Stokes parameters, 5 wavelength points, and a signal-to-noise ratio of 1000) applying ME inversions to any of the Fe I lines at 525 nm. We also find that telescope diffraction has important effects on the spectra coming from very high resolution observations of inhomogeneous atmospheres. Diffration reduces the amplitude of the polarization signals and changes the asymmetry of the Stokes profiles. Conclusions: The two Fe I lines at 525 nm meet the scientific requirements of IMaX, but Fe I 525.02 nm is to be preferred because it leads to smaller uncertainties in the retrieved parameters and offers a better detectability of the weakest (linear) polarization signals prevailing in the quiet Sun.