I do think the best lower risk, greatest capability is to make a generation II Collins. Essentially the same hull, new diesels (japanese?), new battery (lithium? or atleast some improved lead acid or dual lead and lithium banks), electric mast. As a consequence of the new battery tech you would have loads more room in the hull to stick in systems, more battery, vertical launch TLAM etc. You will get greater performance, range, reliability etc. You get an incremental improvement on what we already have (best conventional?), direct swap for existing Collins (allowing decommissioning and hopefully simple IOC).Maybe the first batch should simply be the best that can be designed and built at this point of time (perhaps an “Improved Collins”) with more advanced requirements being designed into a follow on class where time is not such an enemy.
Tas
Build 6 of these now. The remaining 6 could be a new build clean sheet. Essentially a new clean sheet hull. By the time you finish building those 6 generation 3, your already building the replacement for the generation 2 (with the generation 4). You minimise your changes, the old hull proves your new propulsion or systems reducing risk for the new hull.
I look at how things are done in the commercial world with things like Intels (tick tock) or Holdens (prove the powertrain in the older body) or Falcon/911 (continuous evolution) and wonder why it doesn't happen more often with Military projects.