Improve handling of bytes data

[villebro]

This is a continuation of #6987 which caused a regression when selecting non-utf-8 bytes data on Postgres. Also improves and harmonizes rendering of utf-8 compliant bytes data when possible.

Before

• Postgres utf-8 bytes data: 🐍 shows up as 🐍
• Postgres non-utf-8 bytes data: åäöÅÄÖ encoded as latin-1 throws exception
• Non-Postgres utf-8 bytes data: 🐍 shows up as b'\\xf0\\x9f\\x90\\x8d'
• Non-Postgres non-utf-8 bytes data: åäöÅÄÖ encoded as latin-1 shows up as b'\\xe5\\xe4\\xf6\\xc5\\xc4\\xd6'

After

• Postgres utf-8 bytes data: 🐍 shows up as 🐍
• Postgres non-utf-8 bytes data: åäöÅÄÖ encoded as latin-1 shows up as [bytes]
• Non-Postgres utf-8 bytes data: 🐍 shows up as 🐍
• Non-Postgres non-utf-8 bytes data: åäöÅÄÖ encoded as latin-1 shows up as [bytes]

Ping @mistercrunch @john-bodley @mmuru

[codecov-io]

Codecov Report

Merging #7062 into master will not change coverage.
The diff coverage is 33.33%.

@@           Coverage Diff           @@
##           master    #7062   +/-   ##
=======================================
  Coverage   64.46%   64.46%           
=======================================
  Files         421      421           
  Lines       20537    20537           
  Branches     2247     2247           
=======================================
  Hits        13240    13240           
  Misses       7170     7170           
  Partials      127      127

Impacted Files	Coverage Δ
superset/utils/core.py	88.22% <33.33%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c7ffdd6...02ae43c. Read the comment docs.

[mmuru]

@villebro: I verified the changes and it worked in preview and run query modes. 👍