Attributes encoding compatibility between backends

What happened:

Let's create an Zarr dataset with some "less common" dtype and fill value, open it with Xarray and save the dataset as NetCDF:

import xarray as xr
import zarr

g = zarr.group()
g.create('arr', shape=3, fill_value='z', dtype='<U1')
g['arr'].attrs['_ARRAY_DIMENSIONS'] = ('dim_1')

# -- without masking fill values
ds = xr.open_zarr(g.store, mask_and_scale=False)

ds.arr.attrs   # returns {'_FillValue': 'z'}

# error: netCDF4 does not yet support setting a fill value for variable-length strings
ds.to_netcdf('test.nc')

# -- with masking fill values
ds2 = xr.open_zarr(g.store, mask_and_scale=True)

# returns a dict that includes item _FillValue': 'z'
ds2.arr.encoding

# same error than above
ds2.to_netcdf('out2.nc')

What you expected to happen:

Seamless conversion (read/write) from one backend to another. Is there anything we could do to improve the case shown here above, and maybe other cases like the one described in #5223?

Environment:

Reproduce the issue using the shown zarr.group, xr.open_zarr, and Dataset.to_netcdf calls, with and without mask_and_scale. Trace how the Zarr _FillValue reaches attrs or encoding and is passed to the NetCDF backend; done means the demonstrated variable-length string dataset can be converted without the fill-value error.