According to their PR people that is apparently what they did.
This timeline of events suggests that the second DDOS (or "a significant increase in traffic") occurred at 11:46am local time.
At 11:50am local time they blocked all international traffic. This somehow lead to a "short system outage" (which I assume means the whole thing collapsed).
At 4:58pm there was another increase in traffic, "automatically defended by network fire walls". One must assume then that this was all local traffic if we assume that all international traffic was blocked - so either local DDOS impact, or, maybe, new demand from legitimate users.
At 7:30pm though is where things get interesting. There's another "significant" denial of service. This coincides with a lot of legitimate traffic as we enter Australian peak Internet hours. (Again, we can wonder if the DoS was actually just legitimate users smashing their application, but there's no data to decide one way or the other.)
But the fascinating part is that this incident was "significant" because their "geo-blocking service fell over". This apparently then caused a router failure.
First of all, what?! Secondly, from this description it sounds like they were using a server-side geoip mechanism to block the international traffic that was responsible for the DDOS. This will obviously not help in cases where the sheer volume of DDOS traffic is overwhelming the network (which, in Australia, is most of them).
So the question is: was their DDOS mitigation plan limited to simply blocking the DDOS on the server side? Did they not have a contingency to contact their upstream network providers and block entire international routes (which would have cut the impact of most DDOSs off at the knees)?
Sadly most of this information (I think) came from a non-technical press conference, so there's not a lot of hard technical information available yet.
I hope that the ABS will make a lot of their information public - not so that us nerds criticise this whole train wreck (though that will be fun too), but so everyone can learn from the mistakes that were made and we can build better infrastructure.