Avoid metadata requests talking to GCS

[Swatinem]

The GCS Storage would previously fetch the bucket metadata using get_bucket, where a call to bucket is sufficient to create the python class without doing a request.

---

fixes codecov/engineering-team#2276

[Swatinem]

Interesting, I now noticed the metadata updates in a trace for the first time:

[Swatinem]

I updated this with some of the learnings from #351.

This should now be a "minimally invasive" version.

It avoids the needless bucket metadata calls, but otherwise keeps the file metadata call as-is because of the gzip oddity.

I also hand-edited the cassette files to remove the bucket metadata calls from them, and to remove some obsolete ones, but the tests would function even without that update.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 89.11%. Comparing base (d26f241) to head (cfc3232).
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #347      +/-   ##
==========================================
- Coverage   89.78%   89.11%   -0.68%     
==========================================
  Files         375      324      -51     
  Lines       11638     9698    -1940     
  Branches     2063     1735     -328     
==========================================
- Hits        10449     8642    -1807     
+ Misses       1101      992     -109     
+ Partials       88       64      -24     

Flag	Coverage Δ
shared-docker-uploader	?

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.