Just want to say that, on re-reading my statements above, they are a lot more declarative and possibly argumentative sounding than I intended. I don't want to come off as a fanboy, but at the same time Twitter has trained me to treat everything as a debate, even though I'm just disagreeing above, not really actually debating. So, apologies in advance if it came off as harsh.
Lol, all good, did not at all take your response in the negative. Constructive conversations are always welcome.
FWIW, I think one thing that often gets wires crossed in this forum is the timeline of progress; perhaps you'll find that's the case too. So...I agree wholeheartedly with the theory/philosophy of what you're trying to say relative to decreasing cost-to-orbit. But...the reality of today and for many years to come is that those theories are going to stay theories. For instance:
FYI somebody does have the gumption and they are building satellite busses for classes of satellites that don't currently exist, some quite large. They are attacking the %75 BOM cost to reduce that. EG: Much lower cost per pound to orbit means that you don't need to increase the cost of certain components by 10X just to save a few pounds.
Certainly we haven't drawn a box around things like "gumption" and "attacking" so its quite possible we're misaligned there, but unfortunately, my read of the industry doesn't really turn up anyone making strong progress toward a Starlink like BOM for any satellites of consequence. Don't get me wrong--there are a number of entities out there taking sensible, first principals bites out of the cost of an otherwise 'traditional' BOM...but because the first order cost problem is not the cost of mass, there's not a ton of fruit on the tree of "throw mass at the cost problem". The issue is that the first order cost problem is the combination of lifetime, reliability, and performance required by any particular satellite mission.
This is why cubesat companies like Planet and Spire and Tyvak and BCT can build/offer relatively inexpensive satellites...and also why the industry is realizing cubesats aren't actually that awesome when you get to the bottom line of cost. They don't have to last long because can't last long (~2 year missions or so--lots of drag), they don't
have to work because the expectations are low and often they're launched in clusters, and their performance-per-dollar is really not that favorable (and their total performance obviously is way limited) so they're only really useful for a particular type of mission.
As soon as you jump up to the microsat/espa-ish class of sats (Tyvak, Millennium, BCT, Leostella, Surrey...among others) the bottom line price goes way up...but so does reliability, mission life, and available performance.
Where the Starship mass capability really comes into play relative to enabling lower satellite BOM costs is with megaconstelations like Starlink. The important nuance there its that the big enabler is the massive quantity of sats (and ~constant replenishment of said sats) that simply allows the lifetime and reliability requirements to be thrown out the window. To wit, from a traditional perspective, Starlink sats are crap. They have super short lifetimes and terrible reliability. SX would agree, mind, because their math works out that by hucking up a zillion satellites, the statistical crappiness of any one satellite is well mitigated and then some.
Importantly, if not horse beating, that logic only closes with mass quantities of sats. It simply doesn't work for smaller constellations. Someone with tens or hundreds of satellites
can't afford to have a high percentage of them crap out, so they have to pay to make sure they [within statistical reason] don't crap out. AST, for instance, wants to do a couple hundred sats--I think 200 or so. Tyvak's recent Rivada contract is for a 288 sat constellation (and 12 spares). Both of those satellite designs will of course try to leverage down-market components in order to nibble away at the BOM cost, but because their missions will demand near traditional reliability in order to maintain full constellation coverage (let alone service) they're not going to stray to far from traditional qualification philosophies.
