Australia has a perfectly respectable competitive server hosting marketplace and by no means is Telstra a monopoly provider, so even if that justification was true in the (distant) past, it's certainly not true now.
Thelootgnome makes a good point in that we may likely have a localised population. Given there are 12 Oceanic designated servers, and one of the most popular realms, Frostmourne, has about 6,500 characters, I would say it's reasonable to expect at least one fully populated local realm, i.e. 1 player in 12 is prepared to transfer to or re-roll on a local server. In addition, sufficient instance or raid servers would also be required.
If we assumed from the 6,500 characters that equated to approximately 1,000 active accounts, then Blizzard would be looking at $15,000 per month. Which sounds attractive, until you start to question why there has always been such resistance to implement this in the past.
Not knowing how Blizzard's server architecture works, I'd hazard that many of their services would work perfectly well even with the realm and instance servers being locally hosted, but with the remainder of the services being provided out of the US. Services like DNS, identity and access management / directory services, patch distribution servers, promoting code into production and build pipelines and so forth.
However there may still be aspects of their architecture that simply would not currently work over a long distance, for instance connectivity to dedicated storage tuned to WoW's performance requirements, accessing a database cluster over a long distance, intrusion detection or high availability. Sadly the last is probably not a real concern, as we all know from the infamous "World server is down" errors that the realms themselves don't have a fail-over or high availability solution, which can also be problematic over a large geographic separation if this was not an original design characteristic.
Can these problems be solved? Of course they can. But making infrastructure changes of this order is not trivial, nor would be the requirement to duplicate much of the physical infrastructure to support a few local servers to support a new realm and instances. Just thinking through the services architecture work above, my mental cash register easily got into the $1.5M or more price range. Rickman summed it up best that Blizzard has likely resigned themselves to simply not bothering with WoW at this stage of its antiquity, and given its age the WoW architecture almost certainly has many legacy characteristics that make it fragile and difficult to substantially change.
Would there still be hope for Titan? It would depend how much their architecture is a clean sheet design versus re-using their existing services. Needless to say I share the same frustration as all the other Australian subscribers, and in this day and age to lose an entire Tuesday evening most weeks seems a very cavalier treatment of us as valid, loyal, paying customers.