IMO (western) subsurface analysis dance around just how stupid expensive these platforms are relative to opportunity cost of other aquisitions or potential counter measures. Writings will acknowledge nuke boats are pricy multi billion dollar platforms/programs, but don't go the extra step to show the math that each VLS/torpedo tube/cell/unit of fire on a nuke boat costs $100m to hull around, approaching $200m on SSNX, and that's only with more compact TLAMs/cruise missiles... subsonic munitions with very high interception rate - hence gamble on stealth shaping, but doesn't really alter physics that slow munitions are easy to intercept. Extrapolate to more performant but larger prospective hypersonics and you can x2 that multiplier. Factor in usually 1/3 or 1/2 are deployed at any given time (rest training or maintainence) and value proposition gets even more stupid / nonsensical. Then factor in multi week round trip (port and back to theatre) for reloads unless at sea replenishment gets figured out. A B21 costs 700-800m, and can carry much more more ordinances per $ , with significantly greater turnaround. Western nuke boats supremely expensive platform during war per unit of fire, rationalized by increased survivability. But if equation starts biasing toward detection, that rationalization breaks down.
> no magical detection of everything everywhere / AI
We may reach point where where detections/sensors can functionally prevent nuke boats from operating permissively - negating their advantage. Meanwhile quietting has reached returns technical floor. PRC throwing out lots claims of better / cheaper detection methods as they build out their ASW (SQUIDS/ELF waves etc) last few years. Operationally this means better sensors with longer detection range maybe able pickup subs beyond their torpedo range in fleet, or less # of platforms can significantly extend detection bubble/coverage. On the extreme end, you have programs like DARPA ACTUV (I'm sure PRC has their alternative) which relies on dirt cheap surface UVs that just permenantly shadows subs once they leave port, costs $10,000 per day to operate vs $500,000-$1,000,000 for ASW destroyer (US/DARPA estimates, PRC will be lower), i.e. if PRC can push ELF detector ranges, they can basically park ASW platforms right outisde of 24nm contiguous zone off any western sub port (claiming FONAP) and shadow. TLDR is if likely future trend since PRC got into ASW game (relatively recent) is counter to subs is evolving rapidly, likely getting much cheaper, while western subs are getting more expensive, and are multi decade programs that locks in strategic capabilities.
Empahsis/caveate on "western" subs - PRC/PLAN just extended their nuke boat production lines to 4-8 hulls per year. Meaning they see value in nuke boats, but their value proposition per hull =/= USN (or AUKUS). Doubt they'll be paying 6-8 billion per boat.
>If not the submarine, then what
For sea control, there's always PLA rockeforce / air launch AShMs model. 360Billion can buy a lot of mobile IRBM + TELs in hardened shelters to take out shipping (and other targets). 360B doesn't buy a lot of subs, half of which likely will be destroyed in port because they're juicy targets. There's good chance SCS will be so packed with PLA ASW going forward that SSNs will be operating outside of 1IC, in which case subs will stuck in AIP territory off AU coast anyway. There's arguably economical/better/future proof alternatives for shit hits fan. But AUKUS / nuke boats are "good" peacetime procurements, good for photo ops / propaganda / posturing, good for recycling 100s of billions into local economies if AU can actually get the domestic production / supply chain in place. Hence IMO AUKUS will fall through if ends up funnelling 100s of billions into US/UK industry instead of AU (which is where things seem to be moving towards).