Comparison between the observed and modeled EW of the metastable helium line for the near-complete planet sample as compiled in Orell-Miquel et al. (2024). The dotted line is x = y (i.e., complete agreement). The model uncertainties are very large. Rough estimates of the uncertainty due to the XUV flux (based on the scaling in Eq. (3)), the helium abundance (based on additional simulations of TOI-2018 b), and the metallicity (based on simulations of Linssen et al. 2024 of a generic hot Neptune planet) are indicated. An additional significant source of uncertainty is the shape of the stellar SED (specifically the ratio between EUV and NUV flux). Compared to the model uncertainties, observational uncertainties are very small (typically inside the size of the scatter marker) and therefore not shown.