Future components to enhance the basic, native security of 5G networks are either complex mechanisms whose impact in the requiring 5G communications are not considered, or lightweight solutions adapted to ultrareliable low-latency communications (URLLC) but whose security properties remain under discussion. Although different 5G network slices may have different requirements, in general, both visions seem to fall short at provisioning secure URLLC in the future. In this work we address this challenge, by introducing cost-security functions as a method to evaluate the performance and adequacy of most developed and employed non-native enhanced security mechanisms in 5G networks. We categorize those new security components into different groups according to their purpose and deployment scope. We propose to analyze them in the context of existing 5G architectures using two different approaches. First, using model checking techniques, we will evaluate the probability of an attacker to be successful against each security solution. Second, using analytical models, we will analyze the impact of these security mechanisms in terms of delay, throughput consumption, and reliability. Finally, we will combine both approaches using stochastic cost-security functions and the PRISM model checker to create a global picture. Our results are first evidence of how a 5G network that covers and strengthened all security areas through enhanced, dedicated non-native mechanisms could only guarantee secure URLLC with a probability of ∼55%.