The point being that the only thing in the picture that can be considered a unitary object is the mind-moment. Subsequent, related mind moments are not the same thing. They are similar, and they remember being the previous ones, but they aren´t the same person.

I understand, they're not the same "mind-moment" but it doesn't follow that persons are "unitary" objects. Very few things are unitary objects over time. There is nothing incoherent about having words that refer to objects that undergo change.

Because creating a new mind is a positive action. We must be careful about the minds we actually create, because we are responsible for what that mind does, and how it feels. A mind which never exists is unremarkable. If we chose to create one, we should be sure it is the right thing to do.

To create a new "mind moment" I just wait. Even if I care about the distinction between positive and negative acts (which I think isn't a popular distinction here) "not killing myself" does not strike me as a positive action. The word "choice" doesn't seem right here since it isn't like I can reflect on whether or not to "create" a new mind-moment without, in fact, creating one. The word "create" doesn't seem right either. If an object undergoes change we don't say that the object created a new object.

In general, our ethics seem to apply to persons not person-stages. Facts about persons, like that they are collections of psychologically continuous person-stages can generate paradoxes with ethical implications (the existence of simultaneous copies being one). But it doesn't follow that what we're really concerned about is the stages. Take promises. We have a norm that it is wrong to break promises. But if you insist that the person that made the promise doesn't exist anymore (merely because the same mind-moment doesn't exist anymore) the norm collapses. I suspect this is going to be true with nearly all of our ethics because our norms evolved to deal with persons, not anything else.