First of all, Bai swallows whole the myth of "next in line" in GOP nominations contests. Next, I don't really understand what he's trying to say in his glance at history. Yes, Ford/Reagan in 1976 was the last contest that was a "thriller" in the sense of not being decided until the end (actually, the only one like that on the GOP side under the modern nomination system). The nature of these things is that unless a contest is virtually tied, the winner will emerge and overwhelm the field relatively quickly, even if things were very close up to then. I mean, there was, in fact, a lot of uncertainty about the GOP nomination in 2008 for quite some time. It may not qualify as a "thriller," but that doesn't mean that John McCain was "preordained" to win.
Third, Ronald Reagan was in 1976 in no way whatsoever the "Sarah Palin of the day." Reagan, by 1976, was on his second run for the White House, and had completed not one, but two terms as governor of a state that could fit all of Alaska's citizens inside the population of its fourth of fifth largest city (yes, yes, I'm cheating a bit, since California wasn't quite a populous then, but never mind that nitpicking). Not only that, but he had been one of a handful of conservative leaders -- very possibly the single most important one -- for a full decade. Palin, well, isn't.
Mostly, though, Bai gets wrong the entire question of parties and control of nominations. He says:
And even if Mr. Romney or some other candidate were to emerge as the consensus choice of the establishment, this year’s Congressional primaries pretty much showed that the days of anointing are probably over. This isn’t so much a Republican phenomenon as it is the function of an evolving, Web-based society, where your average voter of a certain age isn’t inclined to let his employer or even his church, much less his political party, make his choices for him.This gets things about as wrong as possible. The way to think about political parties isn't to imagine some party establishment, perhaps in Washington, that in the past has dictated policy and candidates to a formerly accepting mass of voters. Instead, parties are collections of groups and individuals, some in Washington and some not, who may or may not agree on all sorts of things. There are activists and politicians, campaign professionals and interest group leaders, members of the partisan press and governing professionals, and even officials and staff of formal party organizations, local and national. When it comes time to nominate a president, all of them coordinate and compete over the various candidates and over the policies those candidates commit to.
Ordinary voters -- voters-as-voters -- aren't involved in that process. Oh, some voters will get more involved, and act as activists; American political parties have always been highly permeable, and they remain so today. In fact, I'd argue that permeable political parties are one of the most important things that makes a policy a democracy.
Voters won't get into the game, in fact, until the Iowa caucuses in early 2012. Even then, hardly any voters show up -- turnout for presidential primaries and especially caucuses is tiny. By then, "the party" -- that is, all of those people a couple of paragraphs up -- will have collectively narrowed the field. More importantly, they will have narrowed the candidates. That is, they will have imposed on them various positions on matters of public policy (just as the leading candidates for the Democratic nomination in 2008 all wound up with more or less the same health care policy).
Now, it's true that sometimes the party has serious internal disagreements about issues. Sometimes those are hashed out and compromised; sometimes they aren't, and leave winners and losers. Thinking of one faction as "establishment" and another as "insurgent" (or even worse, one as party and another as outside of the party) rarely helps understand these fights, however. So, for example, in Kentucky last year the groups and people within the party who supported Rand Paul defeated the groups and people who opposed him, but that doesn't mean that "the party" or "the establishment" lost, since the winners had as solid a place within the party as the losers, and included many long-time Republican actors.
(Is it possible for a true, outside-the-party takeover of a nomination? Yes: see, for example, the cases in which Lyndon LaRouche supporters won Democratic Party primaries. But these things are quite rare).
Sometimes, for whatever reason, all of this doesn't leave the party with a single candidate by the time of the Iowa caucuses. Usually, the next step is the coordination and competition continues. Voters, then, have a say, although even at that point what they do is highly mediated by party actors. The party (and, again, that means all those groups and individuals I mentioned above) also may use primary election results as cues about the popularity and electoral prospects of the remaining candidates.
Of course, that still leaves lots of questions about how exactly the party coordinates, when it fights, and how those fights are resolved. But that's what's happening, not Bai's fantasy that once upon a time "party leaders" picked candidates, while now the voters do.