Add end-to-end benchmarking scripts

[spenczar]

You can invoke these from the root of the repo like:

./scripts/benchmark.sh ./tests/data/index/dataset_500/dataset_500_observations.csv

This will:

• create a precovery database in a temporary directory
• load the data into the database, measuring the time that takes
• do a precovery search on 10 orbits, measuring the time that takes
• print some results to stdout
• write detailed results to two JSON files (precovery_create_benchmark_{timestamp}.json and precovery_search_benchmark_{timestamp}.json) in the current directory

This code was largely written by GPT-4, with some minor manual tweaks.