meta(`benches`): add more benchmarks

[yash-atreya]

Component

Forge

Describe the feature you would like

• Invariant Benchmarks. Ref: perf: fuzz/invariant benchmarks #3411

#3411
set up daily runner of https://github.com/grandizzy/fuzz-benchmarks/ + add more tests
see details in #7610
run #4590
report as suggested in https://github.com/fuzz-evaluator/guidelines

• Isolate: forge t --isolate

• Dynamic Test Linking

• Benchmarks config via toml + allow repo specific config such as setting env variables

Additional context

No response

[0xalpharush]

Here are some codebases we discussed benchmarking in the past:
https://github.com/bgd-labs/aave-v3-origin/tree/v3.3.0/tests/invariants
https://github.com/morpho-org/morpho-blue/tree/main/test/forge/invariant
https://github.com/SorellaLabs/angstrom/tree/main/contracts
https://github.com/ajna-finance/ajna-core/tree/master/tests/forge/invariants
https://github.com/maple-labs/maple-core-v2
https://github.com/sparkdotfi/spark-psm

I think it's useful to disable failures as we are mainly interested in coverage:
#9727

It'd be nice to run with these metrics but disable the corpus and mutations (ablation) to compare coverage guided versus random. Logs can be scraped and plotted over time. This would also allow running just forge test --match-test invariant and not require the line coverage reports (which I believe doesn't support some codebases).

The configuration component is a bit unintuitive because some flags that apply to fuzz tests don't apply to invariant mode e.g. timeout uses an env.,FOUNDRY_INVARIANT_TIMEOUT.