REF: consolidate object dtype handling for any/all

[mzeitlin11]

From #42085 (comment), for groupby any/all we have bool coercing logic in

pandas/pandas/core/groupby/groupby.py

Lines 1521 to 1526 in 6ca8757

	if is_object_dtype(vals):
	# GH#37501: don't raise on pd.NA when skipna=True
	if skipna:
	vals = np.array([bool(x) if not isna(x) else True for x in vals])
	else:
	vals = np.array([bool(x) for x in vals])

Would be nice to share this with handling in

pandas/pandas/core/nanops.py

Lines 534 to 539 in 6ca8757

	values, _, _, _, _ = _get_values(values, skipna, fill_value=True, mask=mask)
	# For object type, all won't necessarily return
	# boolean values (numpy/numpy#4352)
	if is_object_dtype(values):
	values = values.astype(bool)

though complicated by shape requirement differences and the complicated nature of _get_values

[jbrockmendel]

I'm on board in principle, but it isn't obvious to me how to move forward with this. Closing.