I think the main issue with most “full-stack roadmaps” is that they look complete on paper but ignore how people actually learn. They present a long checklist of technologies, but real progress usually comes from building projects and solving problems, not ticking boxes.
A lot of experienced devs point out that learning isn’t linear—you end up picking things up in parallel and going deeper where needed, rather than following a strict path.
Maybe the better way to use roadmaps is as a loose guide, not a rulebook—focus on building something real, and let that project naturally tell you what to learn next.