Ok, let me clear couple things:
1. water freezing - have you ever tried to freeze a bottle of strong alcohol (vodka, whiskey, etc) - you need a temperature far lower than you would experience anywhere in Australia. At about -20 C vodka usually gets a bit "oily" but still liquid, and that's about 35-40% alcohol, rest water (and for those unaware - ethanol is the alcohol we drink in all alcoholic drinks)
2. ethanol absorbing water - out of wiki: "Mixing equal volumes of ethanol and water results in only 1.92 volumes of mixture." So yes, ethanol will absorb some of the water or rather the mixture of alcohol and water will be of smaller volume than a sum of both, but still much bigger than any of them alone. So the described test will still work, as ethanol mixes better with water than with petrol, it will transfer from fuel into the water and you'll be able to see those 2 liquids separated.
3.
If you have water in your fuel that water will show as droplets at the bottom of the tank. Even if you do have water mixed in the fuel and do this test, the water will settle at the bottom of your container raising the initial water level, thus indicating a false alcohol content (if your fuel had no alcohol in it) or will just give you a higher alcohol percentage value if you do have alcohol in it.
Also since E10 fuel have 10% alcohol any water mixed with the fuel shouldn't get too much bearing on the test result itself.
As to the effect of all this on engine performance - I have no idea, but I assume since water won't burn you will effectively have less fuel in your cylinder for the same amount of air, and also you'll have water absorbing some of the explosion energy, so I'd guess your engine power will be reduced.