Software systems constantly evolve to accommodate stakeholders' and environments' requirements that change over time. In this process, the frequent modifications can increase software complexity and negatively impact its global quality when conducted in an unstructured way. The evolutive nature of mobile environments led researchers to investigate how mobile app evolution impacts complexity, security, resource consumption, maintainability, usability, and accessibility. In particular, there has been limited research on the impact of app updates on mobile accessibility: most studies focused on tracking the number of accessibility violations found by automated tools across successive versions of a small set of applications. In a previous work, we made a contribution to this field by identifying accessibility reviews associated with app updates and prompting ChatGPT-4 to provide an overview of the main accessibility issues and enhancements perceived by users in the new releases of a mobile app. In this manuscript, we extend our previous work by adopting manual content analysis to delve deeper into our research questions and by adding new research questions associated with the identification of reviews linked to app updates, user characteristics, WCAG principles and guidelines, and user demands reported in accessibility reviews. Our results show that the accessibility barriers reported by users are mostly linked to the WCAG 2.2 Perceivable principle, and the Distinguishable and Adaptable guidelines, which includes poor color scheme, small font size, unlabeled elements, and lack of customization options. Accordingly, the consequences of the lack of accessibility is mainly connected to the difficult users experience to perceive elements of the interface (e.g. difficult to read and distinguish content, watch videos) and to use screen readers, in addition to feel discriminated against. The most common demand developers and organizations receive is to bring back some accessible feature or to fix accessibility bugs.