Elon Musk tweeted that Starship SN8 failed to land due to insufficient header tank pressure:
Fuel header tank pressure was low during landing burn, causing touchdown velocity to be high & RUD [...]
In this answer, it is explained that header tanks are separate tanks to be used exclusively for landing, to avoid all the problems that would arise from using the partially empty main tanks (ullage, insufficient pressure..). It that answer is correct, I would expect that the header tank is not used during ascent, so it should keep the pressure that it had upon launch (maybe even a little more if it got warmer?). If that's the case, then having insufficient pressure at landing could be easily be solved by pressurizing the header tank more before launch. This answer to SN10 failure instead talks about adding helium as pressurant as a fix. Why would that be needed if the header tank can be simply pressurized with CH4 (or LOX) before launch?
In this other answer about methane header tank having many holes, it is said that the header tank is not separate from the main tank, but it actually gets fuel from it, possible even when the rocket is leaning to one side (during belly dive?). Given that Starship might have to make two consecutive landings on its own (one on Mars, then one back to Earth) with autonomous refuelling in the middle, it might make sense to have the header tanks receive fuel (or LOX) from the main tanks; however, why not make it only when landed and then close the connection while in flight, to ensure that you will have the proper pressure on landing?
