The new solar composition, when applied to compute a model of the Sun, leads to serious disagreement between the predictions of the model and the observations obtained by helioseismology. New measurements of the coronal Ne/O abundance ratio in nearby stars using X-ray spectra typically find high values of Ne/O=0.4 rather than 0.15 normally adopted for the Sun. Drake & Testa (2005) suggest that this high Ne/O ratio is appropriate also for the Sun, which would bring the solar models back in agreement with the helioseismological observations. Here we present arguments why the high Ne/O ratio is unlikely to be applicable to the Sun.