It’s that he thinks the best option is to do nothing. He claims the title of BDFL without playing the role.
See his first reply to the Markdown mailing list in nearly three years: http://six.pairlist.net/pipermail/markdown-discuss/2012-Octo...
He enjoys the credit for being the creator of something used by millions every day, but is entirely unwilling to take the responsibility that comes with that creation being a very public mess.
It works for his usage, but all the ambiguities and undefined behaviors affect a huge number of people, and his only response he's made for eight years has been to retain sole moral authority and refuse to use it.
You gain moral authority by actually doing something. If you use markdown, and wish to propose a better markdown than markdown, go for it, I'm sure lots of people will be pleased, but bear in mind that lots of people will whine about any bugs as well, and expect you to work for them for free and spend significant resources and time fixing edge cases which don't matter to you personally, just because you released the source.
If you're ready for that though, of course go ahead and create a better markdown, or help with this proposed spec, Gruber certainly can't stop you. It doesn't really matter what Gruber says, what trolls on a mailing list say, or what you say on this forum about his responsibilities, whether he is self-appointed dictator for life etc (though I'd dispute that), what counts is putting the work in, which is often surprisingly difficult - far more difficult than criticism.
Right now, the canonical markdown implementation isn't very good. It's hard to change what the canonical implementation is without Gruber's support. It doesn't even require much from him - just his blessing.
He isn't under any obligation to do this. No-one thinks he is. But it would be a good thing for him to do.
If someone wants to write a spec, a better parser or a better markdown full stop, there is absolutely nothing to stop them. The problem is, people would rather bitch about why no one else is doing anything and complain about gruber than actually do the hard work necessary. Do the work first, then look to gruber for canonization if you must, though by that point his views on the matter would be irrelevant.
If the exploits are as well-known as the grandparent comment asserts, and Gruber is aware of them, there really is no excuse for him to leave the code up without any warning that it contains known exploits. However, if he has no idea and everyone is assuming he knows without someone telling him, that's not exactly fair.
Like other responders, I worry that this mentality causes fewer coders to release their projects, for fear of backlash like this post. Think about it: Your feelings toward Gruber are incredibly negative and hostile, and in fact, you would have better feelings toward him if he had kept Markdown to himself and never released it at all. Does that seem fair to you? If the ill will generated by people like yourself outweighs the good will generated by those who appreciate the code I release, or if I fear that it might, what motivation do I have to release my code?