BUG: `Series.__delete__` with EA casts to object

it calls np.delete, which casts to ndarray, which for e.g. PeriodDtype will cast to object

```
dti = pd.date_range("2016-01-01", periods=3, tz="US/Pacific")
ser = pd.Series(dti)

del ser[1]

>>> ser
0    2016-01-01 00:00:00-08:00
2    2016-01-03 00:00:00-08:00
dtype: object
```