Cheaper launches gives defence new use cases too. They'll start to launch smaller satellites into more orbits. It'll take time to change their processes, but they'll be launching small/cube/chip sats with all sorts of sensors.

True but the DoD has a very long planning horizon. So it will take time for them to react.

I can say, buy not attribute, that they are in fact reacting quite quickly.

And for even more context, military spy satellites tend to cost a lot, somewhere in the $billions so way more than the launch. Commercial satellites often share common components and typically cost a few $100 million so not that much more than the cost of the launch. And in the commercial sector the transition to cheaper satellites to match cheaper launchers might be more feasible.

Good point. I hadn't even thought of that. Along with long horizons generally, this probably explains a lot about militaries' "price inflexibility." Launching is a smaller component of more expensive whole.

Same with the big NASA science payloads - if you're spending $8-9B on JWST, paying a $100M premium for a launch vehicle you're slightly more confident in is totally worth it.