> > I see light arrive from X at some time T1+t1
> Except that your notion of T1+t1 is already going to be different from any other (non-stationary) observer's, yes?
Yes, and so is my distance from my current position to X, which I use to calculate my value of t1, which I use to calculate my value of T1. All the numbers are different for all the observers. Special relativity tells us how to convert the numbers between observers.
> > I do a little math to calculate how long it took the light to get there (t1)
> Doesn't this assume that light has a fixed speed? We know already that this ain't necessarily the case, even in a vacuum; there are plenty of things - like sufficiently strong gravity wells - that can slow down light considerably. And in the context of situations involving warp drives and wormholes and other things bending spacetime willy-nilly, that further throws a wrench in attempting to figure out how long it took light to reach you.
First, light does have a fixed speed, locally (that is, at each point within the gravitational well or whatever). Second, gravitational wells are pretty well understood in terms of their effects on light, and I'm pretty sure we could calculate the effect of the warp drive's gravitational distortion on it as well. "The effect exists" does not equal "therefore we can't do the calculation".
> That is: if how an observer is perceiving something is indeed what actually dictates reality
Nobody said that - at least, not until you drag quantum mechanics into the picture, and even then, there are debates about that.
> If causality is to exist at all, even under normal conditions, then it seems like it needs to be possible for observers to be wrong, no?
I have no idea where you got that. If causality is to exist at all, then 1) causal signals can't travel faster than the speed of light, and 2) neither can an observer. If those two hold, then we can have real causality with no observers being wrong, because all observers always observe cause-and-effect happening in the same order.
> > The issue is that I can find a frame of reference where T2 > Okay, which frame of reference would that be? Under what conditions? By what mechanisms that would apply to literally all cases of superluminal communication (even ones explicitly formulated to avoid things like relativistic relative speeds upon which thought experiments like the tachyonic antitelephone depend)?
Not me. I'm not going to walk you through all the equations. I'll tell you where to start, and then I'm done. Start by learning the Lorentz Transform - not just reading the equations, but understanding what they mean and how they work. That will clear up most of your misconceptions. Any basic book on Special Relativity should get you there. The Wikipedia article might, if you take the time to really work through it carefully.