testing: support large arrays

[jdries]

We see tests going OOM when comparing large datasets.
There exists at least one solution that claims to solve this problem:
https://cubed-dev.github.io/cubed/examples/xarray.html

tests/test_benchmarks.py:146: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
venv/lib/python3.11/site-packages/openeo/testing/results.py:517: in assert_job_results_allclose
    issues = _compare_job_results(
venv/lib/python3.11/site-packages/openeo/testing/results.py:449: in _compare_job_results
    issues = _compare_xarray_dataarray(
venv/lib/python3.11/site-packages/openeo/testing/results.py:238: in _compare_xarray_dataarray
    threshold = abs(expected * rtol) + atol
                    ^^^^^^^^^^^^^^^
venv/lib/python3.11/site-packages/xarray/core/_typed_ops.py:551: in __mul__
    return self._binary_op(other, operator.mul)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
venv/lib/python3.11/site-packages/xarray/core/dataarray.py:4817: in _binary_op
    f(self.variable, other_variable_or_arraylike)
venv/lib/python3.11/site-packages/xarray/core/_typed_ops.py:934: in __mul__
    return self._binary_op(other, operator.mul)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = <xarray.Variable (band: 1, y: 68355, x: 33904)> Size: 5GB
array([[[-32768, -32768, ..., -32768, -32768],
        [-327...lValue:           -32768
    scale_factor:         0.01
    add_offset:           0.0
    long_name:            biomass
other = 0.001, f = <built-in function mul>, reflexive = False

    def _binary_op(self, other, f, reflexive=False):
        if isinstance(other, xr.DataTree | xr.DataArray | xr.Dataset):
            return NotImplemented
        if reflexive and issubclass(type(self), type(other)):
            other_data, self_data, dims = _broadcast_compat_data(other, self)
        else:
            self_data, other_data, dims = _broadcast_compat_data(self, other)
        keep_attrs = _get_keep_attrs(default=False)
        attrs = self._attrs if keep_attrs else None
        with np.errstate(all="ignore"):
            new_data = (
>               f(self_data, other_data) if not reflexive else f(other_data, self_data)
                ^^^^^^^^^^^^^^^^^^^^^^^^
            )
E           numpy._core._exceptions._ArrayMemoryError: Unable to allocate 17.3 GiB for an array with shape (1, 68355, 33904) and data type float64

venv/lib/python3.11/site-packages/xarray/core/variable.py:2389: MemoryError