The biggest con to this text is its notation; its insistant to use summation notation for every equation made the entire read an eyesore.
Content-wise, however, it is encyclopedic, exemplified by an impressive bibliography. This is both good and bad. For example, some texts stress that the principle of least action is actually the principle of stationary action. José and Saletan tells us exactly what Lagrangians minimize action...without justification. In fact, because the book is so encyclopedic, a vast amount of material under justified, the most egregious example being Darboux's theorem. Because of this, I am not sure I can recommend this as a stand-alone text for any course. However, as a supplement, it excells.
José and Saletan stress geometrical intuition as a way to decode the equations of dynamics. In this endeavor, the text is a success. I think that because of this, the text makes a great supplement to Arnold's masterpiece, helping the reader visualize the geometrical framework that Arnold uses to construct classical dynamics. Overall, I strongly recommend the text.