feat: add hash_tree_root() bench

[estensen]

First PR in a series to improve the performance of hash_tree_root()

This work will eventually supersede #161 (aligned with the author)

Here I add scaffolding for a Criterion benchmark.

Next I'll add better benchmark data. I'm considering either gettings txs from a random block or randomly generating tx data. Leaning towards random sampling. WDYT?

cargo bench
...
test result: ok. 0 passed; 0 failed; 94 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running benches/merkleization.rs (target/release/deps/merkleization-a419062d8ee9f395)
Benchmarking hash_tree_root: Collecting 100 samples in estimated 5.2192 s (116k iterations
hash_tree_root          time:   [44.761 µs 44.939 µs 45.163 µs]
                        change: [-3.7546% +0.0491% +3.8651%] (p = 0.98 > 0.05)
                        No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
  1 (1.00%) high mild
  6 (6.00%) high severe