Brave’s goal is to randomize the output of your APIs in such a way that we preserve website functionality while still maintaining privacy. So even though the EFF tool reports that you have a very unique value, this value would be different even for you per-site. This is a very effective measure against cross-site tracking. Even for the same site, your screen size (in this case) would change between browser sessions, or if you cleared history, used a Private Window, etc.
A different approach is to report the same value for all users so that users can “hide in the crowd”. The EFF tool is aware of this distinction in anti-tracking approaches, which is why we’re able to achieve the best score possible overall. You’ll also see that for many APIs measured, the test site correctly reports Brave’s output as “randomized by first party domain”. In the case of screen resolution, we can’t report completely randomized values, which is why we add some randomization to perturb the true value.
You can read more here: https://github.com/brave/brave-browser/wiki/Fingerprinting-Protections. Also note that for screen fingerprinting specifically we’re planning some changes to make sure we don’t end up breaking sites. More soon.