We're moving both X and Y, which corresponds to a slope. To move one by numeric increments and the other according to how much computation time has elapsed is, ummm, an interesting design choice.

We're not very near to this number:

>>> x = 9007199254740991.
>>> x
9007199254740991.0
>>> x + 1
9007199254740992.0
>>> x + 2
9007199254740992.0

We're not very near \$2^{53}\$:

>>> x = 9007199254740991.
>>> x
9007199254740991.0
>>> x + 1
9007199254740992.0
>>> x + 2
9007199254740992.0
>>> x + 3
9007199254740994.0

define a function

math

In the Review Context you took a stab at explaining your approach and its motivation. Write a module-level """docstring""" which explores that in more depth.

Given that a picture is worth more than a few words, consider sketching out the approach on a napkin and adding a photo of that to the codebase and to the question. As presented, the what and the why of your computations are not completely clear, nor how the various Magic Numbers dovetail with that. Since we're focused on an approximation, it would be very helpful to talk about error bounds, and about what happens to them on each iteration.

define a function