Very well then, we retcon two last sessions, you didn't left the pub, we will now do the staff hiring process and then we will travel again, and I will roll for random weather and random encounters, we will waste total of 12 hours to make sure all in-universe fiction changes accordingly to the player whims.
This is what you are asking for right now.
No I'm not. I'm not the person who raised the prospect of a two-day adjustment
and who suggested it would actually matter.
You know what I would say to you if I was at your table and you told me the king is killed a day after we talked to him and it effectively doesn't matter if we arrived at 23rd or 25th, this event would only trigger after we arrived? "Choo choo, all aboard the railroad!"
In this scenario that is of
your creation (not mine), you have decided to adjust all the dates, to incorporate two days of personnel hiring. It's established in the fiction that your PC spoke to the King N days after leaving town. If the date of leaving town is adjusted by two days, than N days later is also going to be adjusted by two days.
If you insist that the date on which the King is killed is not going to be adjusted by two days, then you're stuck with contradiction - the King was dead when the PCs spoke to him! So either we can all pretend that didn't happen,
or the GM can come up with some idea that the PCs really spoke to the King's ghost,
or everyone can agree that the date of the King's death is the day after he spoke to the PCs, just as everyone agreed it was before-hand. I mean, it's not as if the calendar date is much more than somewhat arbitrary colour.
The month after Auril is Eolna, and by the time the player tries for the retcon play has progressed to the in-game date of Eolna 32. All the stuff in the example above - the meeting with the King, the King's subsequent death*, various fallout from both those events, etc. - has already happened in play and been roleplayed through.
The requested retcon that adds two days means the meeting with the King could not have taken place, even though it was played out at the time as having happened. Any resulting fallout from that meeting could also not have happened, meaning a whole lot of play already done at the table just got invalidated.
Does that make my concerns clearer?
* - this death could have been any of
a) a scripted and locked-in plot point, or
b) a known deadline the party had to beat (e.g. the King had been cursed to die at the next full moon), or
c) somehow caused by a different played party (in a multi-party game) which locks in the timing for everyone
I threw in the date change idea as an example of a seemingly-trivial retcon that turns out to be anything but, in order to explain why I wouldn't let it happen.
Why would you adjust dates for some events, and not others, in a way that makes everything confused and incoherent? I don't get it. So no, it's not really any clearer to me what is supposed to be going on here. I mean, is this a thing that has actually happened in your play? The second of the above quotes implies that it hasn't.
In any event, I don't understand why this imaginary fiasco, that comes about because of some weird decision to adjust half the dates in the campaign without adjusting the others, has anything to do with my post that it was ostensibly a response to:
I think the GM of a RPG can handle establishing a few bits of new fiction, and fitting them into the established past facts of the game, easily enough.
I didn't say anything about adjusting timelines, either in whole or in part. I certainly didn't say anything about ret-cons. I talked about adding new bits of fiction - something that happens routinely in RPGing that is anything less than a total railroad.
I don't understand why you think your example of weird mis-play - which has never actually happened? - has any bearing on the issue I was posting about, which is talking about the relative simplicity of doing a thing that GMs do all the time: introduce new NPCs, introduce new elements of setting and situation backstory, etc.