Add benchmarks using triomphe::Arc

[michaelsproul]

This is a companion PR to orium/archery#27.

Using triomphe::Arc in rpds seems to result in some substantial speedups, with no slowdowns compared to std::sync::Arc. The data structures that benefit most (~10-40%) seem to be List, Queue, Vector, and RedBlackTreeMap, with less substantial (2-10%) speed-ups for HashTrieMap. The persistent (non-mut) variants of operations also tend to be more greatly affected.

Full data here:

https://gist.github.com/michaelsproul/00a7a41e2f4c1134ae1f95951c86e617

The benchmarking process is:

cargo bench -- sync # establish a baseline
cargo bench --features triomphe-bench -- sync # rerun with triomphe and compare to baseline

Benchmarks were run on an 8-core M1 Pro CPU.

[orium]

@michaelsproul Sorry, I just renamed the master branch to main. Can you open the PR again after rebasing on main please?

[michaelsproul]

I'll leave this closed as the benchmark functionality is also achieved by #88