[FEAT] Interpret `probability_two_random_records_match` in terms of matches per record

[samnlindsay]

Is your proposal related to a problem?

Intuition around probability_two_random_records_match is limited as it's just a very small number 🤷 Hard to sense check a model when the prior is such an unknown.

Describe the solution you'd like

Some means of expressing probability_two_random_records_match in terms of expected or average matches per record.

probability_two_random_records_match x N records = matches per record

Not sure how to communicate this to a user in a useful way...