Fix numpy comparisons; use == for dtypes and isinstance for arrays

[kounelisagis]

This PR fixes NumPy object comparison issues in dense_array.py and sparse_array.py by implementing two changes:

1. dtype comparisons: Updated to use == operator instead of is, following the NumPy recommendation:

   When checking for a specific data type, use == comparison.
   Do not use is, because data type objects may or may not be singletons.

2. Array type checks: Updated to use isinstance(x, np.ndarray) instead of type(x) is np.ndarray.

These changes resolve errors in Dask and other distributed computing environments where worker processes compare objects across process boundaries, while also improving compatibility with the broader NumPy ecosystem.

Closes #2234

cc. @kylemann16

[kylemann16]

Thank you for taking the time to look at this!

[ihnorton]

LGTM (regression test would be good though, if possible)

[kylemann16]

Looks good to me, thank you!

[kounelisagis]

LGTM (regression test would be good though, if possible)

I thought about it longer than I should have, but I couldn't create one.