Do you have any evidence that one shouldn't? I'm not here to do your homework for you, but a few things that come to mind are proven positive correlation with lifetime earning ability and state measures of wealth and income, negative correlation with prison population percentages by state, and a recent history of improvement in the SAT tests themselves.
The latter is significant, during the 80s and 90s the SATs actually were made less difficult because top schools were not seeing the high scores they once did and did not want to lower their visible thresholds for admission so they successfully applied pressure to the Educational Testing Service to tweak the numbers upwards. In the last decade increasing complaints from all schools about the reduced screening effectiveness of the test led to a restructuring of the tests, for example, testing writing abilities.
Considering the period when the SATs were "dumbed down" so Harvard and others could look good, the 40 year decline in performance described in the articles above is actually worse than the score data would indicate. One might also wonder how much denialism about the validity of the tests impacted student performance, if Joe Know-Nothing's kids picked up the message that they didn't need to perform on no steenking SAT tests because the tests were not meaningful, pressure to perform in the areas the tests measure may have been removed.
You're on an incorrect thread of causality. Success on the SATs has virtually nothing to do with intelligence; it's about half information regurgitation and half test-taking skills. Scores have declined over the years because a fair portion of the public has realized that standardized tests are no measure of future success and has stopped treating them as such. The fact that people who do well on their SATs are more likely to succeed is irrelevant; simply doing well on the SAT has a significant effect on college admissions, scholarships, and the like - enough to skew percentages that the ETS will no doubt use to persuade people of the validity of its test. I'm not disagreeing with the claim that education in the States is getting worse - nor am I agreeing with it - but in any case SAT scores are no valid proof of anything except the willingness of the public to accept far too much at face value.