Hi Pine,
I agree with Oliver and Kerry on the violence vs. drip-drip-drip. Our success at keeping people from editing matches the height of the threshold-to-edit. There's no violence going on, it's just the annoyance tolerance level that the typical volunteer needs to have in order to be able cross into the wikiverse seems to be going up these days. Thanks for posting the link to that SWOT analysis - is there any more information about the context in which it was made? Though you mention that experienced Wikipedians know New Page Patrollers create collateral damage and the SWOT slide shows NPP as both a strength and a weakness, no one ever addresses the fact that NPP also contains some seriously misguided POV pushers who in fact act as censors when they use the AfC tools and leave the entries they don't care about to rot. So I see NPP not just as a strength and a weakness, but also as a threat, and its associated AfC is a major threat. I assume AfC was left out of the SWOT analysis because it is so controversial:
In that slide the parenthetic notes on items which may fall into more than one quadrant do not include "undisclosed conflict-of-interest editing" so I am also guessing whoever made it falls into the camp of thinking that anyone who is a paid editor and doesn't reveal this on their userpage is a threat. It leaves out a major battle ground about whether or not GLAM employees are considered paid editors or not when they venture into our world for the first time. I see AfC as one of the major threats to the English Wikipedia, and one way to combat this is to educate newbies on how to avoid it. Whenever I speak to anyone who is interested in contributing on the English Wikipedia and who also has a paid job or a volunteer position that they want to write about on Wikipedia, I first try to change their mind and try to get them to contribute to another subject first and if not that, then another project first such as Commons or Wikidata. I also always tell them 1) never edit Wikipedia from their work IP address in case someone links their onwiki interests to their job and 2) never attempt anything on English Wikipedia through AfC as it will rot there forever. I recommend they ignore the official advice to indicate their employer on their userpage, and always emphasize that their choice of username should be personal and not organization-based. In the case of volunteers this is really hard to explain and unfortunately most have to experience some form of onwiki-harassment before they get it.
I personally see AfC as some sort of last-gasp effort to keep the hoards at the gate. If anything, it enforces a form of old-boys-network onboarding as people have no hope of having their contributions rescued without the help of an insider. I am an experienced Wikimedian of several projects and can wrangle my way through the most complicated templates, but I give up entirely when it comes to the jungle of AfC. Are there any numbers available on how many AfC's resulted in articles vs. submissions and then also the same numbers for non-AfC page creations vs. speedy deletions? Post creation period analysis (3, 6 and 12 months later) would be interesting to have too. My gut feeling is that AfC is busily creating a backlog we will never get through.
To answer your original question, I think the reason we lack New Page Patrollers for the subjects in which they are needed at AfC is exactly the same reason why those people are being turned back at the gates - nobody currently cares about their contributions. If an AfC comes in for a soccer player, ship model or popular TV episode they are welcomed with open arms. You always get what you reward most in the end.