We should initially order 8 or 9 submarines, provide money for the first three, let some future government take over the funding for the others. Remember by the time the 9th submarine is being built it will be late in the 2020's at least.
I like the figure nine or twelve. Built in lots of three, with each three modernised over the previous three.
Based off prior numbers, I think lots of four make more sense. Many military and naval items follow the rule of three, while subs seem to follow a rule of four.
One potential concern I would have though, is without there being a firm commitment for 8+ subs, if only 3-4 were funded, a future gov't might decide to 'save' money by an early cancellation, resulting in penalties and only 3-4 subs in RAN service. Each sub would also be of 'higher' cost because there are fewer platforms to spread the development and penalty costs across.
Now I definitely understand the concern about more money for subs taking away from funding Army and/or RAAF needs, and vice versa. One of the issues that I have observed over the last decade or so, is the number of times that a defence project is 'forced' to do something to reduce costs, which lowers initial/upfront costs, but raises the price tag further on down the line. And it is not even deferring some of the costs, but actually increasing them, much like a balloon mortgage. The ADF does not seem to be forced to do this to the degree or with the frequency that the NZDF does, but it does seem to be commonly recurring event.
One of the other recurring ogres in the corner, is the degree of politics which drives purchase decisions. This can be anything from industrial interests wining and dining politicians and non-defence decision-makers, to politicians choosing who/where to build or employ facilities, based upon whose seat is threatened.
One of the things I tend to like about US defence procurement programs, is that there tends to be long-term, bipartisan planning involved, so that there is not very many changes or breaks in procurement, even when there are changes in the administration (which happens every 4-8 years).
Using the naval construction industry in Australia as an example, if bipartisan agreement could be reached, the RAN and associated agencies have enough demand for vessels for the industry to sustain itself without a boom-bust cycle. Unfortunately what has been happening is that there have been significant gaps in procurement with changes in gov't, starving the industry of work. Then when things reach a sort of critical mass, where orders need to be placed, the various companies then compete with each other for the work and to restart capabilities they had but had been/were forced to allow to wither.
Having it set up, so that companies know that before their current project is completed there will be a follow-on project of some sort, so that companies can plan on transitioning from project to project instead of closing up shop until the next project starts, which might be in six months, or not for another six years... Doing it like this can help reduce and spread out the sunk industrial costs and time. But it does require bipartisan agreement, otherwise it could end with a chance in gov't.