Currently PolicyManager contains functions with a policyID input which is mirrored in higher level python (agents, actors, etc.) PR #2687 sets the policy ID as the HRAC leaving a naming discrepancy and duplicated length constants. I'd like to revisit the motives behind policy Id construction, and unify the result.

• Is there any advantage to using a different hash digest for HRAC versus Policy ID?
• Perhaps there's an anonymity consideration when involving alice's stamp?
• Do we really need to rename on-chain variables for this purpose?

(old) policyID = hash(bob_stamp + label)
HRAC = hash(alice_stamp, bob_stamp, label)

I suggest simply moving entirely in favor of HRAC for simplicity and unifying the usage across the codebase: HRAC can be used for all kinds of policy identification including on-chain indexing.