Create GH action for MACI benchmarks

[samajammin]

Motivation

As a user and developers I want to know how much time and cost it takes to use MACI. I should be aware of the efficiency and costs (both time and gas cots) that I could incur on when using MACI (primary tasks - deployment, voting, proving).

Details

Once benchmarks have been built (#1194 & #1195), build a GH action that can be run in CI to monitor how changes to code affect the benchmarks

We already have some benchmarks (e.g. merkle tree), we made a GH action for this on every PR. We could tweak that existing GH action to run every 3 days. This will be intensive & may start incurring costs, so let's keep an eye on how nightly runs go.