Both arrays start with the GZIP header (31, 139, 8), which is correct for identifying GZIP-compressed data.
The payload (data after the header) in the second array seems corrupted or invalid. GZIP is a format with a CRC32 checksum and size validation, so even a minor corruption in the payload can result in decompression failures or garbage output.
Both C# and Python fail to decompress the second array correctly, which strongly indicates an issue with the data itself rather than the decompression code.