[NO-TICKET] Add new internal stat for profiler performance and document existing stats

[ivoanjo]

What does this PR do?

This PR introduces a new trigger_sample_extra_sleep internal statistic to the profiler, as well as documenting the most important statistics when observing the cost of profiling samples.

Motivation:

We've been looking at these stats and discussing them, so it seems useful to have a reference for them.

Change log entry

None.

Additional Notes:

N/A

How to test the change?

Since these stats are internal we don't have very strict tests around them on purpose -- they can change at any time.

[pr-commenter]

Benchmarks

Benchmark execution time: 2026-02-10 16:09:45

Comparing candidate commit 8d46f4f in PR branch ivoanjo/improve-profiling-overhead-stats with baseline commit 476256d in branch master.

Found 0 performance improvements and 5 performance regressions! Performance is the same for 39 metrics, 2 unstable metrics.

scenario:error - error tracking, with_error=false - third_party only

• 🟥 throughput [-836.844op/s; -791.590op/s] or [-6.817%; -6.448%]

scenario:error - error tracking, with_error=false - user code only

• 🟥 throughput [-821.651op/s; -776.299op/s] or [-6.700%; -6.330%]

scenario:error - error tracking, with_error=true - all

• 🟥 throughput [-413.003op/s; -388.866op/s] or [-5.829%; -5.489%]

scenario:error - error tracking, with_error=true - user code only

• 🟥 throughput [-391.297op/s; -369.512op/s] or [-5.577%; -5.266%]

scenario:tracing - Propagation - Trace Context

• 🟥 throughput [-3467.765op/s; -3345.961op/s] or [-9.284%; -8.958%]