this exists.

look up "Green OA" (https://openscience.com/green-oa-vs-gold-oa-which-one-to-cho...) or "Institutional Repositories".

the reasons these sites haven't taken over academic publishing are many and varied:

  - they are woefully underfunded (library services are low-status, gendered work, that never receive the funding they deserve).
  - publishing in established journals is singularly the most important thing in academic careers.
  - scholarly societies (IEEE is referenced elsewhere in this thread), publishers, and other entrenched interests have significant influence on policy.
  - etc...

in short, this isn't a technical problem. if it were, it would have been solved over a decade ago.