That feels like a surprisingly weak moat though; costs have already fallen to the point where launch isn't the biggest cost of space hardware any more, the competition is hotting up, and whilst launch costs give Starlink an advantage over other LEO satcomms constellations, other countries have strategic incentives to underwrite the existence of that competition, and once those assets have been sent to space it's a straight fight for subscribers in a remote broadband connectivity market which is definitely real but also looks... actually not that huge, relative to a trillion dollar valuation, unless they're able to drop their prices to wired broadband levels without service degradation. Launch cadence is a bigger advantage for SpaceX than cost, but again something other entities plausibly will match, when the demand is there.

The real question is what comes first: viable commercial large scale infrastructure in space that might create new demand for SpaceX launches, or the competition?

SpaceX is pitching their own orbital data centres as a ready to go source of demand for lots and lots of Starship launches, but the unit economics of those vs boring old ground-based server racks and solar farms look dubious even before one considers just how convenient a justification it is rolling Elon's loss making businesses into the IPO.

Valuations are always more of an art than a science but in what world is SpaceX worth more than Meta today? Maybe the $1.75T is to find that world.