BUG: .max() gives wrong answer for non-nanosecond datetimeindex

[MarcoGorelli]

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

In [3]: DatetimeIndex(['2015-09-22 08:31:00', '2015-09-19 09:02:00','2015-09-19 14:14:00'], dtype='datetime64[s]').as_
   ...: unit('s').max()
Out[3]: Timestamp('1970-01-01 00:00:01')

Issue Description

should give '2015-09-19 14:14:00' Timestamp

Expected Behavior

should give '2015-09-19 14:14:00' Timestamp

Installed Versions

INSTALLED VERSIONS

commit : ad91ead
python : 3.11.1.final.0
python-bits : 64
OS : Linux
OS-release : 5.10.102.1-microsoft-standard-WSL2
Version : #1 SMP Wed Mar 2 00:30:59 UTC 2022
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : en_GB.UTF-8
LOCALE : en_GB.UTF-8

pandas : 2.0.0.dev0+1157.gad91ead6de
numpy : 1.25.0.dev0+363.gbb2769e12
pytz : 2022.7
dateutil : 2.8.2
setuptools : 65.7.0
pip : 22.3.1
Cython : 0.29.33
pytest : 7.2.0
hypothesis : 6.52.1
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : None
IPython : 8.8.0
pandas_datareader: None
bs4 : None
bottleneck : None
brotli : None
fastparquet : None
fsspec : None
gcsfs : None
matplotlib : None
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pyreadstat : None
pyxlsb : None
s3fs : None
scipy : None
snappy : None
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
zstandard : None
tzdata : None
qtpy : None
pyqt5 : None
None

[jbrockmendel]

take

[MarcoGorelli]

@jbrockmendel have you already started this? I'd just started a PR, but happy to let it go if you're already on it

[jbrockmendel]

go for it