Yes, Ambinder's post is clearly insane. People have this tendency to want to read deeper and deeper into every campaigns motives for everything. We saw a fair amount of that in Iowa, with all sorts of people suggesting that Vilsack had ulterior motives for running and then again that he had other reasons than his stated one (money) for dropping out. It's ridiculous.
I don't think your polling justifies your point at all. With Clinton the frontrunner based at least partially on name ID, removing any candidate could/should generally increase her performance. When Democratic primary voters have had an opportunity to get to know more than one or two candidates, these polls could and probably will shift dramatically.
The Perfect Storm, I think, was more about media than about persuasion. Having random Iowans show up at your door is not really any more persuasive than random Californians. I'm sure if Dean hadn't tanked on caucus night the Storm would have been credited as genius organizing. Really though the flaws in the Iowa operation were a lot deeper than what one weekend of canvassing could fix, and all the field in the world couldn't save Dean from his communication problems.
There isn't a flaw in the poll - it's just that people who are strong supporters of one candidate or another will say they definitely won't vote for anyone else, even if it is someone they haven't heard of. "Won't vote" and "unfavorable" are not the same thing by a long shot.
None of these things would satisfy me. Primary debates are about Democrats picking a side, not showing our candidates to a general election public. It will get negative, there will be attacks, and it is just going to make us look bad in front of people who will be gleeful about it.
I heard a lot of remarks about how the lists we had this year were the best our longtime volunteers had ever had. The database synced up with the secretary of state's so that we had almost 100% accurate information on who had and who hadn't turned in their absentee ballots. Beyond that it is more a matter of how effectively volunteers keep track of lists (partially training and partially just diligence) and how well staff coordinate the cleaning of those lists. I am basically a list nazi, so I think we handled that aspect pretty well in my region. I'd guess that list quality varied across the state but was generally better than in 2004.