When asked if the EA application pool was representative of the general applicant pool, Dean Rapelye didn't actually address the question. She said the standards were identical. But objectively, the results are different, so we are left with two hypotheses.
(1) That the EA applicant pool is not really representative. It is self-selecting for people who have parents and/or counselors who know how to play the game, or have the resources to hire someone who does. That poses its own set of issues, but is more palatable than:
(2) That offering EA and admitting half the class early helps Princeton (and all the other super-selective schools) game the USNWR rankings, which weight yield as well as selectivity. I recall there was an effort to drop it a few years back, but the other schools wouldn't follow suit.
I dislike EA as good for the school, bad for the applicant (even though both my daughters were accepted EA at their first choices). When it roughly doubles your chances, it's too important a card not to play. So it discourages trying for a reach when you can improve your odds so much at a less-selective second choice with EA. Also, it favors students who don't have to wait to compare financial packages, although Princeton's no-loan aid policy mitigates that.
Everybody back into the April pool, I say.