Benchmarking CLI improvements

[bobbinth]

A few small things that would be great to improve in benchmarking CLI:

1. We should use structop instead of clap as thats what we use for all other CLIs (e.g., here).
2. We should keep CLI parameters as consistent as possible across different CLIs (e.g., for how we specify security, number of outputs etc.). See example here.
3. Let's change the executable file name to something like miden-bench (miden-benchmarking-cli is way too long).
4. miden::prove() also includes execution time, I believe. So, to make things more clear, we should probably label it as "Execution + proving time" or something like that.
5. Would also be great to print out proof size.

[Dominik1999]

Should we revive the CLI benchmarking? Did you use this or are you mostly using the benchmarking of the examples in the Miden VM?

[bobbinth]

I think it would be good to do so - but fairly low priority. So, maybe we can delegate this to someone external?