A couple of days ago I was on the verge of taking this matter very seriously.
Then I looked at the net growth rate of the Wikipedia page count. Around 1500 new pages per day remaining undeleted after seven days. I compared it to the sclerotic maximum capacity of Articles for Deletion (AfD) (average of 112 listings per day early June to early September) and the fact that AfD participants complain bitterly about their inability to keep up with AfD at present.
Bottom line: AfD doesn't scale. Whatever problems may exist with AfD's tendency to randomly delete a selection of perfectly reasonable articles, and VFU's growing unwillingness to rectify this, the problem will decrease in significance in the long term as page creation rates accelerate inexorably beyond the reach of AfD, and possibly even any defensible extension of speedy deletion.
Perhaps requiring editors to sit a short examination on deletion policy would help to keep inappropriate nominations down and increase the overall capacity of AfD, but this would not address the problem of scale.
We should speedily delete obvious rubbish, and we do. In time this manual mechanism may be augmented by cooperative RC (recent changes) monitoring tools, which would improve our efficiency.
But AfD is already swamped by growth and this will only get worse. Correcting its strong exclusionist bias and restoring its connection with the deletion policy would not solve its fundamental brokenness.