The problem I have with full-stack engineers is that they are - usually - rubbish champions for the user.
True, proper, UX&D has unfortunately fallen by the wayside in many technical products, and it is a difficult and extremely skilled specialisation in its own right, and its practitioners should be respected and treated as such.
The absence of real strong UX&D skillsets on software products have directly led to the ‘Enshitification’ of technical products, and made just about every NORMAL person HATE systems that are seemingly designed (or mis-designed) to frustrate them! This is not good for the industry, this is not good for profits, and this is especially not good for society.
The engineers working on the server-side.. due to how they have had to spend their lives in order to achieve real competency in their skillsets.. likely have SUCH high familiarity with using computers, and such advanced IT skills.. that they will NEVER truly understand the position of the average user!
Also, really amazing interfaces are an artistic endeavour - all the well-praised Apple innovations have essentially been good examples of ‘design art’. So you have to be an artist, at least a graphical one. You cannot expect most engineers to be competent graphical artists!
Front-end interfaces should be LED by UX&D, and front-end developers should essentially be UX&D experts working on the interface ONLY and feeding the more technical problems and concerns to the server-side engineers.
In the age of AI this has only become more evident, as AI is truly POOR at championing the needs, requirements, lived experience and psychology of users of software systems - simply because it will never really understand them! It also has zero concept of visual beauty and the experience of being human and interacting with systems. This includes accessibility.. user testing.. user demographic research.. all these things are IMPOSSIBLE for GenAI to understand.
Therefore - the ideal team, in my view, involves TWO types of engineers - UX&D heavy front-end engineers, and AI-assisted server-side engineers (a combination of Test, Cyber, Devops, ML/Data Science and Development skillsets).