[drudman]

You're wrong, I'm right.

It does matter, from a quantum point of view, that you observe. The past does not occur, that is to say all of the wave equations remain an large set of possible outcomes, until they are observed. Google Schrodinger's Cat, or Quantum Suicide Machine.

The second refutation you made is also wrong. Many-worlds hypothesis makes it so that you don't go back in time along the path you came, which you seem to imply. Branching of paths does not occur along a linear time.

"Going back" in time is a misnomer, you don't go backwards on the time line, you travel to a different world that is at an earlier time. This removes the problem of quantum determination, because nothing has been determined.

If time is linear, and many-worlds is either not true, or only good in the future direction, then time travel is impossible. So either it's bunk, and the whole discussion is moot, or it's not bunk, and you're wrong. Choose your fate.