Though you should note, SAT prep has a negligible effect on scores. Intelligence, sadly, is mostly the result of genetics and non-shared environment, with an emphases on genetics. Our inability to admit this would be hilarious if its results weren't so tragic.
Raising the ceiling would also help with the Ivy-league's insane penchant for passing on some of the most able students. Scott Aaronson had perfect SAT scores at 15 - this was back when the SAT was much harder. In his words:
I admit that my views on this matter might be colored by my strange (though as I’ve learned, not at all unique) experience, of getting rejected from almost every “top” college in the United States, and then, ten years later, getting recruited for faculty jobs by the very same institutions that had rejected me as a teenager [...] I was a narrow, linear, A-to-B thinker who lacked depth and emotional intelligence: the exact opposite of what Harvard and Princeton were looking for in every way.
With a higher ceiling, sufficiently high scores are impressive enough on their own. I don't care about those institutions, it's Princeton and Harvard's loss. And Cornell's gain! But it's an insane way to do admissions and it has a real cost on kids. I have nightmares about promising programmers and mathematicians being forced to practice piano instead of doing what they love.
My point would be that there will inevitably be proxy metrics for entry into the elite, and they will tend to be flawed in a way that advantages current spawn of the elite. Once a sufficient mass of people possessing "true" merit knows to chase after those proxies, they will change again to maintain the current composition of the upper class.
Suppose student A has talent 100, and student B has talent 90. Student A puts in a normal level of effort, and scores 100. Student B dedicates all of his time to studying, goes to cram school, etc., and also scores 100. Now they look the same, but A's higher talent of 100 is still going to give them an edge on the job. If you're trying to figure out the candidate's talent, what do you do? You have to look beyond test scores.
Basically, at some point putting too much effort into it is a way of gaming the system, since you're optimizing a proxy instead of what people are really looking for; and changing or expanding the proxy actually makes selection more accurate.