BUG: Series[object].unstack preserve object dtype

[jbrockmendel]

• closes #xxxx
• tests added / passed
• Ensure all linting tests pass, see here for how to run them
• whatsnew entry

Sits on top of #44593

[jreback]

looks good if u can add a whatsnew

[TomAugspurger]

@jbrockmendel can you confirm whether the slowdown in https://pandas.pydata.org/speed/pandas/#reshape.ReshapeExtensionDtype.time_transpose?python=3.8&Cython=0.29.24&p-dtype='datetime64%5Bns%2C%20US%2FPacific%5D'&commits=8374c656 pointing to this is expected? I think it is, but wanted to confirm.

[jbrockmendel]

can you confirm whether the slowdown in https://pandas.pydata.org/speed/pandas/#reshape.ReshapeExtensionDtype.time_transpose?python=3.8&Cython=0.29.24&p-dtype='datetime64%5Bns%2C%20US%2FPacific%5D'&commits=8374c656 pointing to this is expected? I think it is, but wanted to confirm.

No. I don't see how this would affect DataFrame.transpose. More plausibly #44593 which this sits on top of? Even then I'd need to take a look bc I did not expect that to hurt perf.

[jbrockmendel]

OK I think #44593 caused us to split a 2D DatetimeArray into column arrays, which would account for the slowdown, but is not intended. Will look into a fix.

[jbrockmendel]

Perf regression fixed in #44724