Based on the information, Dr. Germain's test appears to have proved that high-school seniors' scores on the test and high scores in their freshman gpas one year later are two events that tend to happen together (75 out of every 100 times). This is the furthest conclusion that can be interpreted from a correlation coefficent.
But correlation does not imply causality, so he cannot predict with these results that those who got senior scores in high school will perform as good (and better than others) when they become freshmen.