The paper by A. A. Shashkin et al. [cond-mat/0409100] reports measurements of the thermodynamic magnetization of two-dimensional electrons in silicon. Although the experimental data is very similar to that reported by us more then two years ago, the authors arrive at an opposite conclusion regarding the spin susceptibility "critical behavior" and spin instability in the vicinity of the metal-insulator transition. We show that this interpretation is based on a flawed analysis of the experimental data.