Early last month, CJR published an interview with Hans Noel, a Georgetown University political scientist and co-author of The Party Decides: Presidential Nominations Before and After Reform. The “key insight” of the book, in Noel’s words, “is to look at presidential nominations not from the point of the view of the people trying to get the nomination, but from the point of view of the party that’s trying to bestow it.” So what does that mean for campaign reporters? According to Noel:

One thing you could do is… you could have someone be responsible for learning about what’s going on in Iowa. So they would go and talk to the various party leaders in Iowa, various activists, people who have been influential in earlier campaigns. You would cover Iowa, rather than covering Michele Bachmann in Iowa. It’s daunting to say, go and understand a whole state. It’s harder than it is to follow around a particular candidate. But that is the place where the questions need to be asked.

Via Seth Masket, I see that the Huffington Post is now taking on that challenge, in a unique effort that combines the political and methodological acumen of the Pollster.com crew—whose talents HuffPost acquired last summer—and the local reporting footprint of Patch.com. (Patch is owned by AOL, which bought HuffPost in February.) The site yesterday published the results of its first survey of GOP leaders in Iowa, New Hampshire, and South Carolina. In that initial article, HuffPost’s Mark Blumenthal explained the project:

This first report involves what will be an ongoing effort by our Patch local editors to identify and regularly query a wide swathe of Republican party leaders, local elected officials and campaign activists about the ongoing presidential contest. We are building the list from the ground up, with local editors identifying these influential Republicans with old-fashioned reporting and then asking them to answer questions via email. We are starting in Iowa, New Hampshire and South Carolina, and plan to both grow the number of participants in those states and expand to other key primary states.

We are calling our participants “Power Outsiders” because these are the people outside-the-beltway who have local influence—which means real power—as the election heats up. You probably won’t recognize their names, but their opinions will likely have great influence on the outcome of caucuses, primaries and elections at the precinct, county and state levels.

While we will ask standardized poll questions, these results cannot be considered a scientific sampling of a larger population. Yet they will amount to the most extensive and systematic effort of its kind to monitor the opinions of Republican activists and party leaders in the early primary and caucus states. As our project expands, we’ll poll influential members of both parties on hot political topics and on issues being ignored in Washington—such as the unfolding unemployment crisis.

This sounds like a great experiment, and congratulations to HuffPost for trying it. As Masket notes, the questions in the initial survey are pretty limited, and the results were hardly earth-shattering—the “Power Outsiders” felt that Michele Bachmann, who won the Ames straw poll, and Rick Perry, who formally entered the race, did the most to help their campaigns over the past week. (Noel, in good social-scientist fashion, told HuffPost that the fact that the “outsiders” views align with the conventional political wisdom doesn’t resolve the question of which way the causality runs.) But as the horse race unfolds, the survey may provide a chance to see the intra-party discussion unfold in something close to real time—and a useful metric against which to measure the national narrative.

I have a few further thoughts, some on coverage of the “invisible primary” and some on how this project fits into political journalism more broadly.

On covering the primary:

● While the empirical approach behind this project is exciting, anytime numbers are involved, there’s a temptation to assign too much significance to them. Sometimes, the “invisible primary” results in a clear winner. But, as Jay Cost noted Wednesday in a smart piece for The Weekly Standard, its function is in many ways to winnow the field, eliminating unacceptable and non-viable candidates and framing a choice for voters once the primaries and caucuses begin. At this point, it seems clear that Mitt Romney and (probably) Perry will survive the winnowing. What we can hope to learn by talking to local party leaders is not necessarily who will win, but just how strong Perry is and whether or not Bachmann belongs in the top tier of contenders.

● Another caution against over-interpretation: local party leaders obviously have, as Blumenthal writes, “real power” in the race for the nomination. But just as obviously, they’re not the only ones with power. Elected officials in Washington, Republican governors, major party funders, and party-aligned media will all eventually weigh in as well. The HuffPost project is valuable because it’s focused on a previously under-covered part of the party’s conversation about who the nominee will be —but it’s still only part of the conversation.

● The value of HuffPost’s experiment really depends on how effective Patch reporters are in identifying important party actors. That’s real on-the-ground reporting, and it’s hard work—because, as Jonathan Bernstein notes, that group isn’t just the “establishment,” and neither is it stable; people are always competing for influence within the party. In an impressive display of transparency, HuffPost has published the names of every person who’s agreed to participate in their surveys. Are there any local political bloggers who want to weigh in on how good these lists are?

On the journalism landscape:

● Though its orientation is different, this effort is in some ways an analogue to the OffTheBus citizen journalism project HuffPost launched during the 2008 campaign, which is returning for 2012. Both projects are innovations in campaign coverage that emphasize what’s happening at the local level—an approach that plays to HuffPost’s mass, even populist, appeal, and plays down the comparative shallowness of its inside-the-Beltway connections.

● This point may be both too banal and too grandiose, but an experiment like this could never have happened in the pre-Internet era. The combination of national readership and robust local reporting resources that’s required might have been available at the major newsweeklies, or some newspaper chains, when staffing levels were at their peak. But the political media’s interest in research-based empirical and analytical approaches didn’t materialize until political scientists and like-minded observers started publishing their own work for a popular readership. Once it was clear an audience existed, some major outlets joined in—as HuffPost did when buying Pollster.com, or The New York Times did in acquiring FiveThirtyEight, where Nate Silver has just put up a great post of his own that is, in part, about how to understand the invisible primary.

We do a fair bit of worrying at CJR about how revolutions in the media world distort the incentives facing journalists. This project is a small example of the corresponding benefits—of how the web’s openness creates not just opportunities for different forms of stories, but stories based on different ways of thinking.

Greg Marx is a CJR staff writer. Follow him on Twitter @gregamarx.