>There is no comparison between the detailed, auditable financials that we have for a project like SLS, and the one or two headline numbers that SpaceX has told us about something like Falcon 9.
No one is claiming otherwise, which is why the WSJ article based on said confidential internal SpaceX financials is so interesting.
As others keep telling you, what matters to NASA, DOD, and other SpaceX customers is what they pay per launch, not how much SpaceX's costs are per launch. Obviously, if a launch cost DOD $50 million but it cost SpaceX 9000 megazillion dollars, there would be some concern on the customer's side about whether such a launch could be purchasable again later, but a) we've had many, many dozens of launches that SpaceX has successfully pulled off, so the internal cost per launch for SpaceX can't be that many megazillion dollars. b) According to the WSJ article, SpaceX has achieved profitability as of this year, so we can now be fairly sure that the cost is, in fact, less than $50 million per launch.
Your statement that SpaceX may be booking Starlink launches as revenue is irrelevant. Hollywood accounting works because a film studio can, by shifting expenses to a particular film from elsewhere in the company, always claim that the film is unprofitable; the overall revenue and profit/loss for the studio is unaffected. The documents WSJ viewed are for SpaceX as a whole, not "SpaceX's launcher division". Even if the launcher division charged the Starlink division 9000 megazillion dollars per launch, that would be exactly balanced by the Starlink division having to book 9000 megazillion dollars of expense per launch. Again, according to the documents WSJ viewed, SpaceX overall has achieved profitability. That's a major accomplishment given it being by far the lowest-cost launch provider in the world.