Remove static_input from data pipeline and model call signatures

[frodre]

This PR finalizes the removal of the StaticInput handling by the data pipeline. The passing of static_input objects are removed from the data configuration, batch iteration, and model call signatures in favor of the direct model handling introduced in the previous downscaling PR (#954).

Changes:

• add get_fine_coords_for_batch to facilitate translation of an input batch domain to output coordinates via the models stored information. For now, this relies on the model's static_inputs, but will be switched to model's stored coordinates in (Add fine coordinates to the model for easier inference handling #971)

• inference Downscaler now takes the batch input_shape instead of static_inputs to check the domain size and model type (regular DiffusionModel or PatchPredictor

• downscaling torch.datasets generators for BatchData no longer include StaticInputs

• removed _apply_patch and _generate_from_patches from StaticInputs

• config.py no longer references static inputs as an argument

• Tests added

[frodre]

Lazy initialization, like done with inference.py to wait for availability on the output domain information

[AnnaKwa]

LGTM, just a couple small suggestions

[AnnaKwa]

Can remove null_generator from utils now

[AnnaKwa]

Suggestion: take the ClosedInterval and LatLonCoordinates as args rather than the BatchData to remove the dependence on how BatchData stores this info (e.g. won't have to change this code if switching BatchData to store the coords as LatLonCoordinates instead of BatchedLatLonCoordinates)

[frodre]

I'm going to leave the single argument method rather than expand into 3 arguments since I think the consolidated passing works in our favor. Expanding would force the repeat of the access pattern that wherever the method is used (and then we'd have to make those updates anyways in multiple locations when we to LatLonCoordinates at the batch level).