This post from a diary on the DailyKos site (and the many comments added by readers) is the most interesting post-mortem I have yet read on the Democrats' failure to win the election.

Susan, the post's author (no last name given), was one of the Dean campaign's California state house party coordinators, and in the course of being evaluated for a similar top position in the Kerry campaign, she was told, " "To be blunt, this is a fat-cat top-down campaign. The campaign staff doesn't really get grassroots."

She then supplies chapter and verse on how the Kerry campaign indeed failed to get grassroots organizing, and then, for good measure, argues that the reliance on 527s also badly damaged the Democratic effort because these groups were legally prohibited from telling voters why they should vote for Kerry, only why they shouldn't vote for Bush.

Dig into the comments and you will find several pungent attacks on MoveOn.org for being too top-down (including some interesting hints of a movement among its members to demand some accountability from its leadership), harsh recriminations against Bob Shrum for being better at winning internal turf wars than the battle against Bush (including the idea that grass-roots Democrats punish him by withholding support from future candidates that hire his firm!), and some fascinating scuttlebut on the role of various unions in the 527s. Yes, you have to wade through a fair amount of verbal chaff to get to the kernels of wisdom, but there's a ton of good, raw intelligence here.

Why is this important? It used to be that after a campaign was over, the principals would write an evaluation that, depending on the campaign's outcome, might never get circulated or only be seen by a handful of insiders and top funders.

Now, the participants--whose intimate knowledge of events on the ground gives them credibility even if they don't have the "god's eye view" of the whole political battlefield--are involved in writing the campaign evaluation too.

This should throw shivers down the spine of every insider who has profited from the clubby insularity of the consulting business. But it ought to be welcomed by everyone else.

Some questions that remain: Is Kos's open process too chaotic to help make sense of all these competing impressions? Would a wiki work better? Is there critical mass to carry this process into evaluating state- and local-level political campaigns, or is the presidential race a noteworthy exception?

And when will the chief players feel compelled to answer and be genuinely held accountable for the decisions they made on behalf of so many others? They can run, but they can't hide.