It's a little unclear exactly what the scores even mean, Consumer Reports doesn't share their methodology. They're just unitless numbers supposed to represent the relative likelihood of needing a repair. But the sense of the description is that these are supposed to represent the risk for a new car, not an existing one.
Really... this seems like a study designed to get clicks and not make real predictions. An actual result with units saying something like "16% of vehicles surveyed lost more than 1 day of utility due to maintenance problems in their first year" would be really interesting. No one has that, alas. So we all end up arguing about bad science.
(My car is just great, FWIW.)