Safety applications are among the key drivers in VANET research. Their study is complex as it encompasses different disciplines, from wireless networking to car dynamics, to drivers' behavior, not to mention the economic and legal aspects. This work presents a simulative study of emergency braking applications tackled by embedding a mobility, cars' dynamic, and driver's behavior model into a detailed networking simulator (ns-3). The results, derived both at the network and at the application level, capture correctly the interactions of the communications and protocols with the car's adaptive cruise control system and the driver's behavior for cars that are not equipped with communication devices. The paper presents in detail the improvements we contribute in simulation techniques and model completeness. It introduces a novel and easy message aggregation technique to empower message re-propagation while controlling the network load during the peak due to the emergency braking. Finally it discusses the effectiveness of such applications as a function of the market penetration rate, showing that even cars that are not equipped with communication devices benefit from the smoother and earlier reaction of those cars that can communicate and whose adaptive cruise control implements a correct deceleration strategy.