On IP fire walls, and other factors determining what actually goes into your new submarine.
Imagine you have an ideal state of an ultimate dream team system of systems submarine design. It uses the very best individual components, no matter where they come from, the concept being it has the very best sensors, hull, weapons, propulsion and auxiliaries available, all operating seamlessly with the very best ships operating system and combat system.
Sounds great but then you need to start applying fixed factors such as the government had specified that the new submarine will be conventionally powered and will use the current USN combat system, USN spec coms and electronic surveillance systems. This automatically entails power and heat budget issues on the design that simply are not a problem for the USN as their boats have such a surplus of power generation they can run virtually everything simultaneously, including such powerful and effective air-conditioning and cooling systems they need to install heaters to make their subs habitable. DEs are very different, especially operating in the tropics where their maximum cooling capacity is governed buy surrounding water temperature, run too many systems concurrently and you cook the boat systems and crew as well as drain the batteries / exceed to output of you AIP.
Then there is the issue of the various systems being scaled to suit your application without compromising their performance, it can work out better to compromise on your selection of the "best" in favour of perfectly good enough and it fits than to compromise on using something that's performance has been degraded to make it fit.
Then there is the espionage issue. Nation states and competing companies spy on each other, it's a simple fact. That spying can take the form of code with in the operating software of a piece of equipment, or even can be hard wired into a the control hardware. This could even, in extremes be an active, or offensive element, able to impair, damage or destroy ship systems, but is more usually used to collect data.
The US has legitimate concerns about integrating competing OS systems with their cutting edge systems. Its not just competitive advantage but operational security at risk, imagine how much China would pay for insights to how the USNs subs do things. It need not even be technically stolen data provided for sale but more likely competing systems, incorporating features and capabilities gleaned from the Australian integration project, an awful lot can be determined from interfaces.
To prevent this from happening there can be no direct integration, rather there must be a firewall or an overarching system that both the US and other systems plug into that prevents the flow of any data from the US to the other systems. Both add complexity and cost, both add inefficiencies and prevent full, seamless integration not just of different elements of the CS but of the CS to the platform. A French or German hull has French or German diesels, motors, electrical systems, auxiliaries etc. each with their own control systems, cabinets, racks, cards, processors, controllers and software, so many different places intelligence gathering could covertly take place. Imagine a condition monitoring system on one of the systems that was over looked that passively collects data from any and all systems it is cross connected to. It is a nightmare.