`NTuple{UInt8}` not getting correctly written out

similar to #411 but corresponds to `fixedsizelist.jl`:
```julia
julia> data1 = (; x = [(0x01, 0x02), (0x03, 0x04)])

julia> Arrow.write("/tmp/julia1.feather", data1)

julia> data2 = (; x = [b"\x01\x02", b"\x03\x04"])

julia> Arrow.write("/tmp/julia2.feather", data2)

julia> data3 = (; x = [(0x0001, 0x0002), (0x0003, 0x0004)])

julia> Arrow.write("/tmp/julia3.feather", data3)
```

```python
In [12]: pyarrow.feather.read_table("/tmp/julia1.feather")["x"]
Out[12]:

[
[
0102,
0304
]
]

In [13]: pyarrow.feather.read_table("/tmp/julia2.feather")["x"]
Out[13]:

[
[
0102,
0304
]
]

In [14]: pyarrow.feather.read_table("/tmp/julia3.feather")["x"]
Out[14]:

[
[
[
1,
2
],
[
3,
4
]
]
]
```

Start with fixedsizelist.jl and reproduce the three Julia Arrow.write examples from the issue, then compare the resulting Feather values with the shown pyarrow output. Trace how NTuple{UInt8} and related tuple types are encoded, and consider the issue done when the affected tuple data is written with the intended structure and the examples produce matching output.

Written by the indexing model from the issue text.