This is a whole can of worms, and my response will be biased and untrustworthy, but here's my take:
In the early-to-mid 1990s, serious security research was intensely cliquish. There wasn't a norm of published vulnerability research; in fact, there was the opposite norm: CERT, the well-known public resource, diligently stripped details about vulnerabilities (beyond where to find the patches) out of announcements, and discussions about how vulnerabilities actually worked was relegated to "secret" lists like Core, which were of course ultimately leaked to became BBS tfiles.
Ranum came to prominence in that era. In the mid-to-late 1990s, after Bugtraq took over, there remained a sort of informal best friends club of, like, Ranum and Dan Farmer and Wietse Venema and like one or two young vulnerability researchers --- Elias "Aleph One" Levy, for instance. There was a sort of acceptance of the idea that Elias and Mudge were doing vulnerability research that was well-intentioned and OK... but that everyone else was just trading exploits on #hack.
There was a sort of focused beam of hatred on eEye, a security vendor that came to prominence in the early 2000s, and most especially during the "Summer of Worms", some of which worms were based on vulnerabilities that eEye's team --- at the time truly one of the most influential teams in all of vulnerability research --- had published. I worked at the industry's first commercial vulnerability research team and had a soft spot for eEye, which was doing the work we did but like several levels better than us, and it has always pissed me off how Ranum and Schneier tried to make hay by dunking on eEye and casting them as "hackers".
(Of course, if you tried to make that argument now you'd sound like a clown, so you won't see people like Ranum and Schneier saying that kind of stuff. But the fact is, the arguments were clownish and inappropriate back then, too.)
So, if you ask me, the argument Ranum is advancing is literally that public vulnerability research is bad, and that details of vulnerabilities should be kept between vendors and a few anointed 3rd party researchers. Because otherwise, you were just helping people break into computers.
The Ranum of 2005 is I think especially characteristic of what I'd call "moralizing" information security; that security is actually a fight between good and evil, that what's important about machines getting owned up is that somebody's livelihood depends on that machine running, and the hacking is a crime, and the details of how the hack worked are about as relevant as the details of how a burglar breaks into the window of a house they're burgling without setting off the alarms or whatever. I get it, but I'm from the opposing school of information security, which is that security is just a super interesting computer science problem.