Main advantage of jest would be execution speed - no browser means fast execution. But also lower confidence as bugs that only show up in a browser rendering engine will never happen in a node test, so can’t be caught.
I’d argue having a consistent DX between e2e and component tests is worth a ton - no context switching, same commands, even shared custom commands all works great. You even get videos and screenshot assets from CI, because it really rendered.
And finally debugging and building in a real browser removes a whole category of pain and workarounds related to fighting jsdom, missing browser APIs, and debugging HTML in the terminal trying to imagine how the component might have looked during the test.
I have post about it on the Cypress blog:
https://www.cypress.io/blog/2022/12/15/cypress-component-testing-for-developers/
And you can find more if you check out the Component Testing tag there.
There’s probably one other win for jest besides execution speed, which is shallow mounting. It doesn’t make much sense to shallow mount in the browser. Users will never interact with a shallow mounted component. But sometimes people are used to doing it and it’s an adjustment.