I'd love for this to be repeated with a piece of text that is not that. A friend who works as an English teacher at a community college pointed out that there might be a bit of cognitive bottlenecking going on - the inability to parse the text declines very quickly once you go above a certain percentage of missing vocabulary.

This comment on there also adds some extra nuance:

I'm inclined to be very sympathetic to the students here. Those paragraphs may as well be in a different language, they're filled with words no-one uses anymore, or which are being used in ways these kids have never seen before (if you haven't encountered "whiskers" or "wonderful" in these contexts before you won't think to look them up in the dictionary just in case) and it's entirely reasonable for college students not to have learned the words for 19th century phenomena they will never encounter in their own lives (horse blinkers, Michaelmas). Even the fact they had a dictionary isn't the catch-all excuse you want it to be because if they have no incentive to get these questions right they won't be motivated to do twenty minutes of linguistic archaeology to answer these questions.

I fully expect the underlying point, that college student literacy has collapsed, is true, but I think the people who designed this test failed to build it in a way that it could possibly prove or disprove their hypotheses.

At this level of difficulty, reading a text like this is a test of subject matter knowledge and reasoning ability, not literacy or English language skills.

I used to get poor marks in French listening, speaking, and writing, but ace reading comprehension, because I had general knowledge and the ability to reason well with incomplete information. It wasn't reflective of my French skills when I overperformed in one of four French tests.