How exactly are you defining "railroading" here? Because it sounds like you're using it differently than I understand it.
My understanding of railroading is that no matter the choice, the outcome is exactly the same. That is not the case here for the simple fact that there were two possible outcomes. One outcome was that Taylor survives, and what comes after will be seen in the next few chapters. The other outcome was that Taylor dies, and from there come the changes of being reborn, dealing with the impact of how that took place, who finds out about it, likely creating a new Guardian Beast, maybe springing a surprise attack on the enemies who are actually your allies; a very different direction than the previous. That, to me, is giving the players agency, just that there is not an option of defeating these opponents at this time.
Any situation which is 'inevitable' is usually seen as railroading if it is the GM who has set it forth.
To put it this way, if there was no reasonable way for players to avoid triggering event 'x', that's railroading. It doesn't matter that 'x' can have multiple outcomes, if players are forced into that event, that in and of itself is railroading. It's basically used to give players direction at the start or whenever they get lost... but doing it mid 'other quest' is something that's really hard to do right without angering people or just not making sense.
It's fine if the players have willingly stumbled into the unwinnable fight, or if it's the start of a story/quest... and the players can SEE how their actions influenced the aftermath even if losing was guaranteed because they still feel like they've accomplished something. This fight was far more like a cutscene than it was something players could've done something about, and Gameplay/Story segregation problems IMO is something that quests are supposed to avoid. Here, again, it's a binary lose-capture / lose-die situation where there was no chance or action apparently for us to create dialogue, or get help or any other outcome or modifier to the 'you lose result'. And I cannot see how it is the players fault for getting into this situation. I cannot go 'well, in hindsight blowing off all these people, and making X/Y/Z decisions caused this' and thus I see it as railroading.
Having the story abruptly take this turn feels a lot like we were being railroaded into a scenario
^ Being railroaded into a scenario with pre-planned results of 'lose-live, and lose-die' is railroading. If there was a reasonable way to avoid the scenario altogether perhaps...
This scene isn't the good kind of screw ball that we can look back and past updates and go 'ah I see how this came about' this is the bad kind where, like in old point + clicks, we needed to think in Moon-Logic and do something complete different than a reasonable/usual manner to achieve something.
We can't see logically how we ended up here from the previous updates. The information we players have cannot lead us to the conclusion that 'the surprise ambush was TSAB and our usual pattern of attack would have them treat us as some sort of super brainwash army robot'.
Honestly? Because combat scenes are tiring and stressful for me to write, so if I can wrap a fight up in one chapter and move on to a different scene I prefer to do so. It made sense to me for this fight to end when it did.
And clearly I misjudged how many hurt feelings there would be about it.
This is the part that is perhaps, the most worrying thing for this quest's future. This is both Worm and MGLN, where combat is a fairly important part of the narrative to both show character growth/resolve and characters themselves. Especially when most of our growth in terms of 'power' revolves around being better in combat situations (and how we're rewarded for getting in fights), if you struggle to write combat... it's probably a better idea (unless you're trying to improve on that) to figure out ways to incentivize avoiding combat scenes or figure out how to handle it more mechanically. If writing combat scenes causes this kind of problem, it is better to figure out some other way to handle it then, perhaps an outline and getting a co-gm to write them instead then?
Thinking it through, the best way to do a death/resurrection in a quest would probably be to hold onto the post for a longer time and then post both the death and the waking up update at the same time. It would present the whole thing as a fait accompli, as sort of a "things largely beyond your control have happened, here's the new situation, what do you do to respond to these events?" It would also give a productive outlet for thread discussion by making people look to the future rather than only having the death to analyse.
^
This too, to comment from my experience with quests where any 'part 1 update' where the Protag is getting hammered and ends on a 'well looks like you're screwed' note without any method for players to interact with (Especially if their plan is clearly failing less than half way through the update, and had their been an INTERRUPT! to allow them to try a different plan that might've done something different) has caused some heavy flak. Especially when you give players several days to stew over it.
I agree with this. If this was the result then it's not a good idea to force the players to to sit on the bad end result for a few weeks. It only really serves to breed discontent - well, more discontent if we're talking about this thread specifically - and bad feelings in general.
Every single AGG quest where something like this has happened has led to literal pages of complaints and flak. Every single time it's happened.