Studies of macroevolution have classically rejected the notion that large-scale divergence patterns can be explained through populational, microevolutionary models. For morphology, this consensus partly derives from the inability of quantitative genetics models to correctly predict the behavior of evolutionary processes at the scale of millions of years. Developmental studies (evo-devo) have been proposed to reconcile micro and macroevolution. However, there has been little progress in establishing a formal framework to apply evo-devo models of phenotypic diversification. Here, we reframe this issue by asking if using evo-devo models to quantify biological variation can improve the explanatory power of comparative models, thus helping us bridge the gap between micro- and macroevolution. We test this prediction by evaluating the evolution of primate lower molars in a comprehensive dataset densely sampled across living and extinct taxa. Our results suggest that biologically-informed morphospaces alongside quantitative genetics models allow a seamless transition between the micro and macro scales, while biologically uninformed spaces do not. We show that the adaptive landscape for primate teeth is corridor-like, with changes in morphology within the corridor being nearly neutral. Overall, our framework provides a basis for integrating evo-devo into the modern synthesis, allowing an operational way to evaluate the ultimate causes of macroevolution.