test: benchmark scope_apply_to_event for events and transactions

[jpnurmi]

The benchmarked payloads are dummy events and transactions with arbitrarily chosen 10 vs. 256 frames and 100 vs. 1000 spans. The exact numbers are not important, but to see that the numbers don't accumulate when the payload grows.

Related to #2042 (comment) by @limbonaut:

Performance suggestion: since this is used to retrieve trace context, we could add a specialized accessor that avoids cloning the transaction tree. On my system, with 20 completed spans on the bound transaction, I measured a ~14% slowdown vs master, and it scales with the number of spans. (I measured the time it takes to capture each individual event over 10k samples.)

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 74.70%. Comparing base (5e2a38d) to head (5ada6e9).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #2100      +/-   ##
==========================================
- Coverage   74.73%   74.70%   -0.03%     
==========================================
  Files         103      103              
  Lines       27386    27386              
  Branches     4944     4944              
==========================================
- Hits        20466    20459       -7     
- Misses       5570     5578       +8     
+ Partials     1350     1349       -1     

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.