Why does a test fail when using findBy but succeed when using waitfor?

Using react-testing-library, the following test works:

it( 'renders popular search terms, with links to search urls', () => {
      waitFor(() => {
        const popularSearch = screen.getByText( Copywriting.buyer.shop.popularSearch, {}, { timeout: 5000 })
        expect( popularSearch ).toBeInTheDocument()

        const popularSearchPills = screen.findAllByTestId( 'pill' )
        expect( popularSearchPills.length ).toBeGreaterThan( 0 )

        const hrefs = popularSearchPills.map( pill => pill.href )
        expect( hrefs.filter( urlIsValidAndNotBaseUrl )).not.toHaveLength( 0 )
      })
    })

But the following test used to work, but now fails:

it( 'renders popular search terms, with links to search urls', async () => {
      const popularSearch = await screen.findByText( Copywriting.buyer.shop.popularSearch )
      expect( popularSearch ).toBeInTheDocument()

      const popularSearchPills = await screen.findAllByTestId( 'pill' )
      expect( popularSearchPills.length ).toBeGreaterThan( 0 )

      const hrefs = popularSearchPills.map( pill => pill.href )
      expect( hrefs.filter( urlIsValidAndNotBaseUrl )).not.toHaveLength( 0 )
    })

Why would the two code snippets function differently? I thought findby was supposed to be a wrapper for waitfor.

Note: what's happening under the hood of the rendered component is that we dispatch an action which calls a saga, the saga calls fetch, which returns a piece of data, the saga then calls another action with the data as a payload, triggering a reducer that saves the data to the store. The view should then update to include the element with Copywriting.buyer.shop.popularSearch.

Is there any reason, on principle, why the two tests should have different outputs?