fix(waterdata): Drop dataclass slots=True for Python 3.9 compat

``slots=True`` for ``@dataclass`` requires Python 3.10. The package
declares ``requires-python = ">=3.9"`` and CI tests 3.9, so the import
was failing test collection on the 3.9 matrix cell. Dropping the kwarg
loses a small memory optimization on short-lived ``_Axis`` instances
(not material) and restores compatibility.

Also aligns one residual "sub-chunk" comment to "chunk" — the rest of
the file already uses "chunk".

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Author: thodson-usgs

dataretrieval/waterdata/chunking.py

@@ -452,7 +452,7 @@ def _request_bytes(req: requests.PreparedRequest) -> int:
     return len(req.url) + body_len
 
 
-@dataclass(frozen=True, slots=True)
+@dataclass(frozen=True)
 class _Axis:
     """
     A single chunkable axis of one user-level request — a list of
@@ -932,7 +932,7 @@ def _combine_chunk_frames(frames: list[pd.DataFrame]) -> pd.DataFrame:
     Dedup is restricted to rows whose ``id`` is non-null. ``pandas``
     treats NaN==NaN as a duplicate for ``drop_duplicates``, so a
     blanket call would collapse every id-less row into a single one —
-    silent data loss if any sub-chunk emits features without an
+    silent data loss if any chunk emits features without an
     ``id`` field.
     """
     non_empty = [f for f in frames if not f.empty]