This paper aims to contribute to the understanding of the main specificities of latecomers' processes of technological development. Building on the basis of this understanding, it searches for the reasons why the conventional measures of Science and Technology (S&T) policies, usually inspired by the so-called linear model and by neoclassical economics, are frequently inappropriate or insufficient for developing economies. This is so because latecomers compete primarily by imitating, rather than by innovating. Such feature of their technological processes compels them to rely on cheap labor, on state protection or natural resources depletion as a way to compensate for its relatively low initial productivity. The cases of four latecomers - Brazil, Mexico, South Korea and Taiwan - are briefly analyzed. During the last two decades, all these four countries were successful in greatly increasing their shares of world scientific productions. The two Asian economies were also able to achieve very large increases in their shares of world technological productions, greatly shrinking the productivity and per capita income gaps that separate them from the levels of the leading industrial economy, the US. The two Latin-American economies, however, went in the opposite direction on those respects. Such fact put into question the linear model's assumption of a more or less direct connection between a country's scientific achievements and its technological production or innovation performance. The paper is concluded by presenting some implications of its analysis for S&T policies for developing economies.