Trump could also only win the presidency without a popular-vote majority because a large region of the country, the greater Rust Belt and Appalachia, had been neglected by both parties’ policies over the preceding decades, leading to a slow-building social crisis that the national press only really noticed because of Trump’s political success. In this sense, Clinton’s weird post-election boast that her half of the country was way more economically dynamic indicated the advantages of a system where a declining region can punch above its popular-vote weight — because it makes it harder for a party associated with economic winners to simply write the losers off.
However: This defense of occasional countermajoritarian presidencies assumes that the political system will, over the medium-term, be responsive to the Electoral College’s incentives — that parties will be capable of overcoming polarization and addressing specific regional grievances, that politicians will be capable of working toward Rooseveltian or Reaganesque majorities, that presidents who win with a popular-vote minority will either adapt and gain a majority the next time (as George W. Bush did) or lose like Benjamin Harrison and John Quincy Adams.
And neither political party has responded to 2016 the way my defense of the Electoral College predicts they should. A countermajoritarian outcome has not produced supermajoritarian ambitions.