I'll pick 0, 1, 5, and

6) Starlink v3 is itself new and nearly untested. Launching a lot of v3.00 models, then discovering that they really need to be upgraded to v3.01 to fix various hardware issues, is not a win.

>6) Starlink v3 is itself new and nearly untested. Launching a lot of v3.00 models, then discovering that they really need to be upgraded to v3.01 to fix various hardware issues, is not a win.

Yes, and that could also tie back into #4:

>Starlink v3 manufacturing is delayed and they don't have enough units, which I find doubtful given Starship is years late and they should have a backlog of v3s waiting to launch

This assumes that a delay isn't completely purposeful. Technology and their experience with Starlink has continued to advance over those years of delay, it's not like they simply stopped iterating with Starlink v2, they just couldn't do everything or put all the pieces together in one package due to size/mass. But I'd be unsurprised if they took advantage of those extra years to keep working on it. They've proven they can ramp Starlink production easily, it's not clear they'd have any reason to produce some big backlog and have it sitting in a warehouse. Seems better to tie it closely to Starship's dev cycle so they end up launching current stuff, not sats that have been sitting around and are now years old already. They can ramp v3 Starlink production alongside Starship.