My initial post on Scott Brown drew a pretty good reaction, and this response from Michael Barone is a must-read. Barone argues if Republicans are able to seriously contest Massachusetts, and in a race where Democrats are the heavy sentimental favorites after the passing of Ted Kennedy, that could have a kind of terrorism effect on wobbly House Democrats that will force many more to retire, and inspire strong Republican candidates to jump into the race in blue districts. This could set the stage in a very nice way for November.
Absent a scientific poll of the race, two recent bits of analysis suggest we may be headed for a close race on January 19th.
The first is the "citizens' poll" being undertaken by Republican activists frustrated by the lack of recent polling. This shows Scott Brown ahead, and the methodology consists of randomly scouring the phone book in select areas. From the photo below, it's a hearty effort, but definitely homegrown:

The results through New Year's Eve were Brown 241, Coakley 216, Undecided 63. A further 201 refused and 213 were left pro-Brown phone messages. That works out to a slight Brown lead of 44.6 to 41.5 percent among those who responded.
But that comes with a big caveat: the areas polled were limited to Worchester and Cape Cod, with a little Quincy thrown in. These areas do not look like the rest of Massachusetts. In both Worchester and Barnstable (Cape Cod) counties, Obama defeated McCain by a margin of 56 to 42, or a margin 12 points lower than his statewide performance.
Adjust the margin, and this would put Coakley 9 points ahead statewide.
As a point in Brown's favor, there's no attempt made to screen voters in the survey. Essentially, it's a poll of adults, not registered voters, or even likely voters. In this environment, I imagine a likely voter screen would help Brown to the tune of about 5 points. Anecdotal information from their responses also suggests Brown voters are way more fired up about turning out:
Many Brown supporters were enthusiastic and upbeat. Statements included: "What other choice is there 'but' Scott Brown?; "There's a lot riding on this"; "Our country is at stake."; and "God help us if we don't get Scott Brown elected."
To the contrary, Coakley voters seemed down, dull, and disenganged. Is it any wonder not one Coakley voter expressed optimism.
This is valuable information. I wonder what such a poll done of 3-of-5 or better voters on the statewide voter file would yield.
A second attempt to fill the void comes from RCP's Sean Trende, who looks at the composition of the electorate in Virginia and New Jersey compared to 2008, as well as the vote swing from Obama to McDonnell and Christie. If Democrats, Republicans, and Independents swing to the Republican Brown as they swung to Republican gubernatorial candidates in both states, that would make for an exceedingly narrow Coakley win of 51.1 to 48.9 percent. Though we are accustomed to thinking of Virginia and New Jersey as two distinct races, interestingly the final numbers suggest a rising tide lifting all Republican boats. In terms of vote spread, McDonnell performed 24 points better than McCain, and Christie performed 22 points better. A similar swing in Massachusetts would get Brown to within 2 to 4.
I don't know that there is a scenario that would show Scott Brown ahead right now, but there are plenty of scenarios I could see that would show this within the margin. And conditions right now seem optimal: a fired up Republican / Tea Party base that is stirring beneath the surface, and a complacent Democratic Party. The question is what happens if a poll does show the race close? The campaign would then take on a whole new dimension. Will the entire electorate expand in a way that would likely benefit Coakley? Democrats will pull out the stops to nationalize the race, and would play the Kennedy card as they have so many times before. But without a Presidential race on the ballot, which is typically the only thing that can create downballot coattails, will it work? Is there a way that a close race could pull in more marginal Republican voters excited about the possibility about picking up a seat in Massachusetts without bringing out more Democrats, enough to close a 3 point gap? We'll know in 18 days.
Full disclosure: As discussed in my previous post, my company provides some online services to the Brown campaign, but we're not involved in the campaign day to day.