ci: QoL improvements for benchmarks

[ddyurchenko]

What does this PR do?

1. Change threshold for marking a difference in benchmark results as a statistically significant.
2. Enable S3 bucket uploading, since permissions are already configured.
3. Change how PR-commenter acts on a new benchmark result - instead of minimizing a PR comment, replace it, like it is done in the most CI/CD tools and in dd-trace-js.

Motivation

Improve benchmark workflows based on previous requests from #apm-benchmarking-platform channel.

Describe how to test/QA your changes

Successful S3 bucket uploading: https://gitlab.ddbuild.io/DataDog/apm-reliability/dd-trace-go/-/jobs/201494066

Reviewer's Checklist

• If known, an appropriate milestone has been selected; otherwise the Triage milestone is set.
• Changed code has unit tests for its functionality.
• If this interacts with the agent in a new way, a system test has been added.

[pr-commenter]

Benchmarks

Comparing candidate commit 20b7553 in PR branch ddyurchenko/qol-improvements-benchmarks with baseline commit cf23647 in branch main.

Found 0 performance improvements and 0 performance regressions! Performance is the same for 6 cases.

[ddyurchenko]

Hey @ajgajg1134 ,

Can you please review & merge this PR?

I changed 2 things:

1. Now baseline and candidate commits are guaranteed to be in correct order for comparison.
2. Fixed plots in benchmark-analyzer that were distorted due to long labels.

Thanks!