The source is so utterly useless, and has been discredited several times elsewhere on this site, that the quiz is rather pointless. It's based on English learners responding to a survey rather than any actual English ability (which would put countries like Seychelles and Mauritius, and probably Ghana and Liberia, near the top).
You're right, at the time when I had made this quiz, I wasn't aware of the deficits of the source nor had I seen it elsewhere on the site. There are clearly huge methodological flaws within their study.
So strange that Portuguese-speaking Angola would have more English proficiency than, say, Zambia, Malawi, Lesotho, Liberia... places where English is actually the lingua franca.
Largely due to methodological flaws within the study. It's not a representative sample of each country because it uses volunteer sampling, which greatly skews the data of the study.
Sketchy.