I have not denied it. Nor have I accepted it as I just have not come across enough evidence that traffic bleed exists.
When I wrote 'traffic bleed' or 'significant traffic bleed' it was referring to demonoids "they'll bleed a lot of traffic" in his post I was responding to:
My comment that you're quoting was in reference to the particular case that Tim referred to. On traffic bleed more generally, which is what has been predominantly discussed in this thread - my comments were a bit different, see pertinent parts below:
I've seen people do it first-hand, they automatically type in the address... .com.au. I've done it myself.
Different extension, but the test below (and others I've done) highlighted traffic bleed for me. It happens, the amount varies but it definitely happens.
In any case where the general public would expect a particular extension (E.g. .com or .com.au) and something else is used - there will be some traffic bleed.
The real question is: will it be a little or a lot?
I don't have the magic formula for you Rhythm, I can't give you a % of traffic bleed from a .net.au to a .com.au - as I'm sure you can appreciate every case will be different.
All I know is that traffic bleed does exist and there are plenty of examples that demonstrate this, some of which myself and others have referred to.
You're entitled to your own views as to whether traffic bleed exists, I'm not going to argue the matter till I'm blue in the face!