Re: Mach's point
Mach has decided that Musk's stated intent to colonise Mars is just words to inspire overtime from his employees. Musk's personality and history lend enormous support to that hypothesis. On the other hand, the rate of engine production and the scale of the test facilities at McGregor and Boca Chica are wildly excessive for developing a rocket that is intended to mostly launch Starlinks. This is based on the enormous number of third party cameras covering the sites 24/7, satellite photography and roughly weekly aerial photography. I could understand the scepticism if all we had to go on was Musk Tweets. (According to Musk's lawyers no reasonable person would consider a Musk Tweet to be a source of factual information.)
Sending a Starship to Mars or the Moon requires propellant transfer. Even sending a large payload to GTO requires some propellant transfer. Starlinks do not require propellant transfer. Mach sees the number of propellant transfer launches required to send Starship to the Moon as a deal breaker. IIRC the number in SpaceX's HLS proposal was 12. Depending on what numbers you pick it could be anywhere between 3 and 19. This is an enormous problem for some people and completely irrelevant to others. From my point of view, the plan is to make Starship launches cheap so a large number does not matter that much to SpaceX. It does not matter at all to NASA: tax payers pay the same whether it takes 3 or 19 launches to refuel a Starship HLS.
If you try to follow Mach's point of view, Starlink does not require both rapid launch cadence _and_ high payload. The full constellation of nearly 40,000 satellites replaced every 5 years only needs a launch every 4.5 days (or more if you reduce the payload below the original target of 150t). Slowing the launch cadence allows more time for propellant boil off, increasing the number of launches required to refill Starshp HLS. Decreasing the payload increases the number of launches. Some combination of the two make Starship HLS and Starship to Mars impractical - if the target launch cost is exceeded by a large enough factor. Mach wants to see Starship as a Starlink launcher that is a bad fit for HLS. He selects evidence accordingly and pooh-poohs anything that counters his point of view. SpaceX's hardware rich incremental development feeds into his obsession. He can point at IFT5 and the likely results for IFT6 as overwhelming evidence that the current version of Starship is completely useless for going to the Moon (no argument there).
The current version of Starship is also useless for launching Starlinks. Lots of space enthusiasts (not necessarily Musk fanatics) expect incremental improvement from SpaceX. There is abundant (non-Tweet) evidence that SpaceX has significant improvements in the pipeline. I have confidence that those improvements will make version 3+ Starships viable for launching Starlinks (from Florida) and that propellant transfer will work at some point - very unlikely to be ready for a landing in 2026. I am sure HLS will reach the Moon. The first one may do so at quite high velocity. (SpaceX does not get a payday for a hard landing.) The next one may fail too - but for a different reason. I am sure Mach will get thoroughly wound up when people like me put "incremental development" in a comment for such an event.