Exactly: any constellation built with pre-falcon9 rockets might be non-Musk, but it won't even be remotely starlink-like. They'd inevitably aim for orbits highe

Exactly: any constellation built with pre-falcon9 rockets might be non-Musk, but it won't even be remotely starlink-like. They'd inevitably aim for orbits higher than "throwaway-LEO" and the much sparser constellations those orbits enable (lines of sight less constrained by horizon) will cause considerably more latency (far from geostationary-bad, but a meaningful quality difference) even if the inherent SNR drawbacks were somehow solved. It seems quite rational for governments to desire a fallback, but it won't ever be more than that and as long as starlink is on the market any hope for significant customer contribution seems unwarranted. Without rocket parity to f9, it's just hopeless to get meaningful customer contribution as long as starlink exists. At least as long as they don't find some miracle tech to massively extend VLEO lifetimes (solar powered VLEO Bussard jet or something similarly far out), but even with that the numbers required would be virtually unachievable without an f9 equivalent (it's a somewhat crazy project even with the f9!). They should absolutely go for it, if they consider having a starlink fallback worth the investment, but they should really not base that decision on illusionary hopes for customer contribution.