Well, as I have said before in the comment thread, I am not printing during runtime. I printed ONCE, to get the number that the physics frames were using as delta at 60 fps, and then slotted that number into the equation. I still get desync without calling any prints.
Furthermore, I can run the print function as much and as randomly as I want DURING runtime, and as long as the replay was recorded with the same motion modifier as I play it back with, I get absolutely 0 desync. The print function is completely irrelevant to this inquiry.