Deserializing list-of-null data leads to incorrect results

[stinodego]

Checks

• I have checked that this issue has not already been reported.
• I have confirmed this bug exists on the latest version of Polars.

Reproducible example

import io
import polars as pl

df = pl.Series([[]], dtype=pl.List(pl.Null)).to_frame()
ser = df.serialize()
result = pl.DataFrame.deserialize(io.StringIO(ser))
print(result)

Log output

shape: (2, 1)
┌────────────┐
│            │
│ ---        │
│ list[null] │
╞════════════╡
│ []         │
│ []         │
└────────────┘

Issue description

An extra row is added for some reason. Must be related to the 'hack' introduced in #17218

Expected behavior

Preserve the original frame.

Installed versions

main