When you are looking at the Australian statistics, you have to remember that the gun related offences are a small component of the total - and when you are dealing with small numbers, any trend in the data is going to be very hard to pick out from the 'noise' in the series.
In the armed robbery stats, the percentage where guns were involved does seem to be decreasing - its gone from 15.5% in 1993 to 6.3% in 2001, or a decrease of more than 50% over that time.
However, year to year variation (those figures hop around like a mad rabbit) is quite high, which would make me cautious of claiming any sort of trend without looking at how the statistics were compiled.
Those graphs are the number of victims, which is not necessarily correlated to the number of offences or incidents - we could just be seeing a trend of offenders going for softer targets (such as the corner store, with one person in it and no security compared to the bank with 25 people, cameras and alarms). I could be wrong, I'm just doing this off the top of my head and I haven't looked at the methodology behind the statistics.
And as for the murder stats - out of some 300 murders, guns are involved in about 20% of them (or some 60 odd murders); this jumps up in 1996 (Port Arthur) but then settles back down to about the same level as before.
If you are trying to say 'gun control works', (as in the Firearms buyback), OR "Nyahh nyahh, its made no difference" you can't really do it from those stats yet.
The firearms buyback was meant to cut down on large scale events like Port Arthur; check the stats in maybe 20-40 years to see if there has been a reduction in gun related mass murders...we don't really get a lot of them, so it takes time to collect enough data to make any conclusions.