Where to start?

My first thought is that you might want to check if it is actually cheaper to build a duplex than it is two separate houses.  If not, and if space permits, why not just build houses?  I would think the value would be higher than for attached dwellings…

You don't necessarily need to split the title…. it is possible that you could keep the existing dwelling, and build another one or two dwellings on the block, and keep all of them.  If you did that, you'd spare yourself  the subdivision fees.  In this scenario, you'd want to look at what the eventual rental return is for the entire lot, and how much you spent to achieve it.  If for instance you worked out that the eventual rental return would be $600 per week, but you had to spend $800k to achieve it, then it isn't really a great return.  But let's say you worked out that you could generate a return of $900 per week for an outlay of $600k, now that's some nice numbers. 

1800m2 with 23m frontages front and back is a very useful dimension.  Depending on council rules and regulations, you'd easily fit four HOUSES on it!  And if you got creative with double storey townhouses, even more.  Which raises an interesting point…. perhaps you don't have to build anything.  Perhaps you could simply apply for permission to do so, and then resell the entire block to a developer for a profit.  Again, the numbers would have to work…

By the way, where on the block is the current dwelling located?  And what is it made from?  Brick?  Weatherboard?  If it is weatherboard and in good condition but annoyingly positioned in the middle, maybe you could "pick it up and shift it over".  If memory serves me correctly, the May edition of Australian Property Investor Magazine had an article on a young couple that use this "slide and divide" strategy well.  In fact they didn't even build new dwellings in the vacant spaces.  They trucked second hand weatherboard houses in from wherever.