Benchmarking code is buggy

[sisuresh]

Running cargo bench --features bench --bench worst_case_linear_models -- --nocapture sometimes fails because the underlying code has changed without any validation being done. Here is an example -

thread 'main' panicked at soroban-env-host/benches/worst_case_linear_models.rs:127:17:
cost type InstantiateWasmFunctions has too large a constant CPU term over InstantiateWasmInstructions: 41007 vs. 28905 (0.41868188894654906 diff)
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
error: bench failed, to rerun pass `-p soroban-env-host --bench worst_case_linear_models`

The goal of this issue is to fix the benchmarks, and then re-run them to sanity check the current settings.