In the process of digital transformation of society, various fields of social and industrial practice move towards the use of semi-automated interactive systems that imply a high cognitive load on the operators who manage them, as well as the users who utilize them. Usability studies lack a conceptual step towards seeing results of task performance as complex cognitive-emotional user states where cognitive load may be diminished or enhanced by the surrounding factors, to the extent that a given user loses functionality. We employ the concept of (dys)functional worker states to explain users’ resulting states after they perform two typical real-world tasks of information search and/or assessing online content. Moreover, formation of complex user states is cumulatively conditioned by four ‘contextual fidelity’ factors, including user traits, task nature, experiment conditions, and interface quality. Thus, we develop a research design that integrates navigation vs. analytical tasks, group vs. individual performance, and monotony- vs. anxiety-inducing conditions. The results provided by 40 assessors and received via measuring five cognitive and emotional metrics show that the navigational task evokes anxiety, while the analytical task induces fatigue and a drop of intellectual lability, both results being simultaneously conditioned by all the three control factors. Partly contradicting previous knowledge on group vs. individual task completion, our results call for multiple-group usability tests for all interfaces, including multi-modal, chatbot, genAI, and industrial ones.