Why can't LLMs be used for long-term care of a software project?

This seems to be more on the tool-wielder than the tool.

Linux itself is receiving LLM-assisted code (a quick search saying 20% at the moment, although I don't know how accurate that is)

LLM use != vibe coding (latter should be reserved to "cattle" projects)

One can use LLMs to find vulns and/or review hand-written code, for example. But to answer your question, one doesn't understand code they didn't design nor write themselves.

LLM don't care. Humans do, but humans with LLMs are more prone to "drive-by" style of software development or substituting "expensive" ingredient (human time and attention) with "cheap" one (LLM tokens).