Almost any mainstream article about computer security portrays the threats as far more sinister than they really are. That's mainly because the whole world of hacking and computer security is exotic in most people's minds and gains an almost-magical aura.
But that being said, the conficker worm is technically interesting and very very well made. ebrown said that it has been "defeated" but he's talking about the exploit that the worm used, which isn't at all what the security researchers are talking about. That hole was closed long ago, they are talking about their inability to take down the bot net that the worm created. The holes are always patched and there are always millions who don't bother applying the patches. That much is normal, what is different is that this network isn't being taken down or rendered useless as easy as large scale infections like this are usually fought.
A "bot net" is a group of computers that have been compromised and that can be controlled by a remote party. These are often used to send spam, infect other computers, or launch denial of service attacks (this kind of attack basically means each computer sends traffic at a target in order to flood them and prevent other legitimate traffic from getting through).
The conficker worm has made a big bot net for someone, but it hasn't really been used, and has been cleverly evolving in ways that make it very difficult to take down the botnet. For example, while many worms will use one specified domain to phone home to and get instructions this worm is setup to try many random domains that are not yet registered. This complicates the efforts to fight it, because with just one domain the domain itself can be taken over. Security researchers have been registering some of the domains the worm will "phone home" to but due to the volume of random domains it can use they can't get them all and this leaves an easy way for the hackers to control the network by just registering one of the domains.
So these kinds of tricks is what makes this threat currently "undefeated", but it is very unlikely to be used for gratuitous attacks. This kind of work usually represents a significant investment, and is worth lots of money to the individuals who control it. Doing something like a DDOS attack to flood a country's internet connection isn't a very useful thing to do with it. It's most likely that this is a worm operated by for-profit hackers. Not a government or for-curiosity or for-politics hackers so those kinds of attacks are unlikely.
So what they are most likely to do with it is find a way to monetize their network of zombie computers. They can sell spamming services with it, they can steal credit card numbers to sell, they can even just "ransom" the infected computers in a process where they sell the fix to the problem they created.
But with all the attention the worm has gotten, some of these methods are hard to get away with. Right now it really seems like they are just trying to preserve control over the bot net and figure out a plan to make money off of it without getting caught.
In short, it's very likely that this will just contribute to more of the spam, fake anti-malware scams, and data theft that is already common place on the web. It's very unlikely that it will be used to do things like attack Australia's internet. If it did something like that, the control of the bot net would be risked for very little gain.
The article is essentially saying that this bot net is more resilient than they have typically been in the past. But the threat is nothing new, just more clever than normal.