This PR: #148 solved a problem of inconsistencies in default browser behaviours when running the e2e for light and dark mode. It done this by forcing all tests to run in light mode.
We should run the e2e tests on dark mode too and assert the differences in colors etc.
