martillo,
The Lorentz transforms relate frames. If you want to relate time in the mothership frame to a particular twin's frame, of course you need the Lorentz transforms that relate those two particular frames.
It's not a matter of opinion. If twin A moves to the right along the x-axis wrt. the mother ship S with relative speed v, i.e. velocity v, and their coordinates coincide on (0,0), the transformation equations relating spacetime coordinates (x<sub>S</sub>,t<sub>S</sub>) from the S frame to spacetime coordinates (x<sub>A</sub>,t<sub>A</sub>) in the frame A are
x<sub>A</sub>=γ(x<sub>S</sub>-vt<sub>S</sub>)
t<sub>A</sub>=γ(t<sub>S</sub>-vx<sub>S</sub>/c<sup>2</sup>)
If twin B moves to the left along the x-axis wrt. the mother ship S with relative speed v, i.e. velocity -v, and their coordinates coincide on (0,0), the transformation equations relating spacetime coordinates (x<sub>S</sub>,t<sub>S</sub>) from the S frame to spacetime coordinates (x<sub>B</sub>,t<sub>B</sub>) in the frame B are
x<sub>B</sub>=γ(x<sub>S</sub>+vt<sub>S</sub>)
t<sub>B</sub>=γ(t<sub>S</sub>+vx<sub>S</sub>/c<sup>2</sup>)
That's the way it is.
What you did amounts to nothing more than observing that two separated events that are simultaneous in the S frame, are not so in the A frame. No surprises there, and certainly no contradictions. This becomes extremely clear when you use frame names as subscripts instead of the ambiguous prime ('). You were transforming from S to A. That doesn't let you conclude anything about B.
So, you were mixing frames. It's a classic mistake, but a mistake nonetheless. Accept it, and move on.