This topic has come up countless times in the 3 years I've been contributing to Wikipedia. Everyone wants some sort of policy which will automagically result in more great articles, but no one wants to do any heavy lifting.
* When we had trouble with poor spelling, someone wanted to deploy spell-check software. But we soon realized that wouldn't work, because of pairs like THEIR and THERE which require artificial intelligence -- which is STILL too unreliable for this.
There is no substitute for good writing. Nothing else will improve a bad article but good writing. Deleting stubs will not *cause* a good replacement article to appear. Votes to determine "consensus" use up 3 to 10 times as much energy as it actually takes to fix a marginal article.
About once a month, someone will VfD an article I've written. A power struggle then ensues. In nearly every case, the outcome is that I improve the article so much that it is not even CLOSE to the originally-listed VfD item -- and the article "survives".
I'd rather people would contact me on my talk page and complain about my poor writing than launch a campaign to condemn my work. Struggling against condemnation is dreary and often discouraging.
My thanks go out to all those who have dropped me a line, saying, "Did you know that the 200-word article you just wrote is on VfD, and the vote is 7-1 for deletion?"
To sum up: I think VfD is a waste of time, and I prefer not to participate in these votes. If an article is VfD'ed, and I think the topic is important enough, I'll just revise it and remove the 'notice' (with a comment in the edit summary or on the talk page explaining my improvements).
Ed Poor, aka "Uncle Ed" Stubborn Pioneer of Excellence in Writing