First, let’s assume, arguendo, that the occasional “second-place winner” is a bad thing. National Popular Vote does not fairly cite the elections of 1876 and 1824 as examples of this phenomenon. The 1876 election was held in the wake of the Civil War and Reconstruction. Fraud and dishonesty pervaded the election. The popular vote totals are not necessarily reliable. Similar problems exist with the 1824 results. At the time, many states’ legislatures were still selecting electors on behalf of citizens. There is no way to know what the popular vote would have been if all states had then relied upon a popular vote as their method of elector selection.
Second, Electoral College opponents often cite “nightmare scenarios” in which the Electoral College produced a “near miss,” just as National Popular Vote does here. As a mathematical matter, the statistics cited may be true. But as a practical matter, a candidate usually could not have gained the extra votes he needed in one state without losing more votes in another state or states. For those who are interested, I recommend reading “The Case Against Direct Election of the President” by Judith Best. Best has completed an analysis of several of these so-called “near misses,” and she concludes that the “[shift-in-votes argument] fails because it abstracts from political realities.”
Finally, I disagree with your contention that second-place winners are inherently bad. In any contest, rules are established for a certain purpose. Let’s consider an example sometimes cited in this context: baseball. Any fan knows that teams do not earn spots in the playoffs by scoring the most runs during the regular season. Instead, they earn their spots by winning the most games in their respective divisions. And they win the World Series itself by winning 4 of 7 games. Naturally, rules could be established to rely on “most runs” instead of “most games won” during this entire process. But such rules would not accomplish the stated objective of the championship game: Allowing the two best overall teams to face off at the end of the year and striving to crown the best of the two teams as the winner. A revised set of rules might allow a team, for instance, to earn a spot in the World Series by having one great month and several poor months. Or perhaps a team that is great at taking advantage of weak opponents (but rather poor at facing off against good opponents) would win a berth in the World Series. Excellent performances throughout the baseball season would not be required to earn the championship. Occasional, stellar performances could be sufficient.
The rules for the presidential election contest are established with a similar purpose. They seek to identify the best national candidate, overall. The system leans in favor of candidates whose strengths play out evenly, rather than those who perform brilliantly in one part of the country but terribly in other regions. Particularly for a country as large and diverse as America, these election rules make a great deal of sense.