IVAS_rend custom loudspeaker output tests limited in inputs

After noticing the bug in #1326 (closed), I inspected the tests and it seems we are limiting the input formats in the custom layout output renderer tests. Here is the part extracted from test_renderer.py

@pytest.mark.parametrize("out_fmt", CUSTOM_LS_TO_TEST)
@pytest.mark.parametrize("in_fmt", OUTPUT_FORMATS)
def test_custom_ls_output(
    record_property,
    props_to_record,
    test_info,
    in_fmt,
    out_fmt,
    get_mld,
    get_mld_lim,
    get_ssnr,
    get_odg,
    get_odg_bin,
    split_comparison,
):
    run_renderer(
        record_property,
        props_to_record,
        test_info,
        in_fmt,
        CUSTOM_LAYOUT_DIR.joinpath(f"{out_fmt}.txt"),
        binary_suffix=EXE_SUFFIX,
        get_mld=get_mld,
        mld_lim=get_mld_lim,
        get_ssnr=get_ssnr,
        get_odg=get_odg,
        get_odg_bin=get_odg_bin,
        split_comparison=split_comparison,
    )

This in practice means that MASA and ISM inputs are not tested for custom loudspeaker rendering. This obviously needs to be fixed unless there is a specific reason for the decision (@kiene @tamarapu).