investigate flaky test

[fxamacker]

Description

Investigate flaky test TestMapIterate/collision

---

• [ ] Targeted PR against main branch
• [ ] Linked to Github issue with discussion and accepted design OR link to spec that describes this work
• [ ] Code follows the standards mentioned here
• [ ] Updated relevant documentation
• [ ] Re-reviewed Files changed in the Github PR explorer
• [ ] Added appropriate labels

[fxamacker]

After updating to the latest commit of stretchr/testify, the timeout hasn't resurfaced after 75 runs (25 x 3 versions of Go on Ubuntu) using the same seed that was previously able to cause one timeout every 3 or 6 runs.

• 25th run using Go 1.15
• 25th run using Go 1.16
• 25th run using Go 1.17

With testify updated, 25 x 6 configs (150 runs) on GitHub ran successfully without a single timeout.

The seed for prng used time.Now().UnixNano().

• CI workflow used a 3x2 matrix of configs: Go 1.15, 1.16, and 1.17 on ubuntu-latest and macos-latest.
• Test logs are at https://github.com/onflow/atree/actions/runs/1472610205
• Prior to this opening this PR, an additional 75 runs successfully completed without a timeout using a hardcoded seed on ubuntu-latest.

However, there's a chance the problem was/is caused by a GitHub issue that has been resolved because the timeout no longer reproduces on GitHub using prior commits that failed every 3-6 runs.