Although it's true that many colleges ignore the SAT essay, but multiple choice portion of the test is *not* highly correlated with academic success. The highest correlation is (sadly) family income, followed by weighted/normalized high-school grades (e.g., not GPA, but a weighted GPA), and only then standardized tests. Also above a certain high score (~1400/1600 on the SAT), there is nearly no correlation at all with higher scores and educational and post-educational outcomes (and yes I used to work with admission committees for a university that cooperated with other highly-selective university to compile statistics on this subject over many years back in the '80s).
The idea that the SAT matters is a myth propagated by the Educational Testing Service (ETS) corporation. In fact the creation of the competing ACT test was prompted by the fact that the SAT origins were an *aptitude* test (that can draw it's lineage from the US army IQ testing recruits in WWI) , not an *achievement* test (testing things that you should learn in school).
Colleges wanted an achievement test, but were dismissed by the ETS, however because of the use of the SAT in ivy league schools, the University of California signed on in 1960 and made the ETS/SAT into a juggernaut. Now because of discontent by UC and other schools on its predictive value, the ETS has changes the SAT twice in 10 years, which in its latest form, now looks more like an *achievement* test (like the ACT was).
Of course there is open debate in higher education on even requiring tests like the SAT or ACT. For example this study tracking 123,000 students over 33 universities found only minimal correlation of academic success with even submitting SAT scores to the school to evaluate (let alone what the score actually was).
