Hi, creator of the site here -- I worked for several years teaching English to foreigners, and my own (informal) statistical research has found that native and non-native language acquisition, while certainly somewhat different, is not tremendously different. I've run this test on Americans and Brazilians at all different levels, and the size of the progression from known to unknown words is rather consistent.
That being said, cognates between languages can give an artificially inflated score. I intentionally avoided any words with the same roots in English and Portuguese (my second language), which should hopefully also be true for most Romance languages. This way, you shouldn't be able to "guess" meanings you've never actually learned. However, it wouldn't surprise me if German speakers, for example, were able to "guess" an additional number of words correctly.
Also, the 60-80 words are only what you are tested on -- the word selection on pages 2-4 changes depending on your answers on the first page, so it attempts to narrow down your vocab knowledge "at the margin".
That being said, cognates between languages can give an artificially inflated score. I intentionally avoided any words with the same roots in English and Portuguese (my second language), which should hopefully also be true for most Romance languages. This way, you shouldn't be able to "guess" meanings you've never actually learned. However, it wouldn't surprise me if German speakers, for example, were able to "guess" an additional number of words correctly.
Also, the 60-80 words are only what you are tested on -- the word selection on pages 2-4 changes depending on your answers on the first page, so it attempts to narrow down your vocab knowledge "at the margin".