Mark DiBarry

Doing some general performance tests before/after. I agree that it's better to go add more stuff and do it the "right way" rather than hacking it with the interpolation data....

Also looks like the culling is correct. CanvasGroup is so good for visualizing it: https://github.com/user-attachments/assets/444b6854-1711-4408-9507-6466b142abd4

Hm. Unfortunately, this seems to break `ParallaxBackground` and `ParallaxLayer`s, which seem to not repeat anymore. We'll have to make sure there's no regressions there.

@kleonc As one solution, we could just have `ParallaxLayer` store data the same as `Parallax2D`. I don't see any side effects from any of my tests with the following changes:...

Also, did some testing with 40 `Parallax2D`s, with 5 repeats each (You shouldn't need it higher than 3, though I know people will use 20 if they can) I was...

> I guess I forgot they exist. 😆 But to be clear you're saying they should repeat the same way as Parallax2D with this PR, not in the pre-this-PR manner?...

If you fix the `repeat_times` issue, I can't find any other problems. LGTM!

@kleonc Oof! I never even imagined anyone would use a parallax effect with y-sorting... and I still can't think of a use case, but I'm sure someone can and will...

@kleonc Thanks for the advice! Made an issue and PR here: #93182