Benchmarks binary compilation takes too long

### Is your feature request related to a problem or challenge?

Follow up on https://github.com/apache/datafusion/pull/18985

I found that building the benchmarks binary takes forever

```
cargo run --release --bin dfbench -- tpcds --iterations 5 --path ../../datafusion-benchmarks/tpcds/data/sf1/ --query_path tpcds --prefer_hash_join true -o tpcds_sf1.json --query 30
```

The compilation takes 5 mins locally, it too much IMO, would be nice to explore if this can be optimized

### Describe the solution you'd like

_No response_

### Describe alternatives you've considered

_No response_

### Additional context

_No response_

Start with the reported `cargo run --release --bin dfbench` command and review the follow-up context from pull request 18985. Investigate why the benchmarks binary takes about five minutes to compile, then verify that compilation time is reduced without changing the benchmark command's behavior.

Written by the indexing model from the issue text.