US/China war scenarios
@Keller: I mean, one could say that eg. linear regression is standard methodology, but you still wouldn't use it on strongly non-linear data, because that wouldn't make sense. If someone submitted a science paper with a graph that looked like this:
https://4.bp.blogspot.com/-KZdQh0ja5hE/VD_AEjp0Z1I/AAAAAAAAHUk/iZqdBRnKXYM/s1600/normal.png
it would get flagged during peer review, no matter how many other papers used linear regressions (I'd hope :) ).
Re importing military components, that becomes very important from a trade or supply chain perspective (vulnerable during embargo, diplomatic crisis, etc.), but I'm not sure how it'd be big enough to matter from a budget perspective. Looking at the US's military budget, as it's more transparent than China's:
https://qph.fs.quoracdn.net/main-qimg-5b376b73caae5bbbda703e65e6f0d035.webp
Only 20% of total military spending goes towards procurement. Of that, a large percentage is going to go towards things simple enough for any major economy to produce fully domestically (food, guns, trucks, body armor, etc.). And then even for a device that might depend on imported parts, eg. fighter jets and missiles and such, a component can be vital while still being a small fraction of the total cost (eg. modern microprocessors for smart weapons). So without precise numbers, it seems that the budgetary fraction of non-replaceable imports would be down in the low single digit percent range.