Defer hashing in staged ledger diff application

[georgeee]

Problem: when performing transaction application, all merkle ledger hashes are recomputed for every account update. This is wasteful for a number of reasons:

1. Transactions normally contain more than one account update and a hash computed for an account update is sometimes overwritten by a subsequent account update
2. Ledger's depth is 35, whereas only around 2^18 records are actually populated. This means that each account update induces an wasteful overhead of at least 17 hashes
3. When an account is touched in a few transactions of the same block, it's gets hashed a few times, whereas in fact only the final hash is truly needed

Solution: defer computation of hashes in mask. When an account is added, it's stacked in a list unhashed_accounts of masks which is processed at the time of next access to hashes.

This fix improved performance of handling 9-account-update transactions by ~60% (measured by #14582 on top of #15979 and #15978).

Explain how you tested your changes:

• [x] Measured performance with #14582

Checklist:

• [x] Dependency versions are unchanged
  • Notify Velocity team if dependencies must change in CI
• [ ] Modified the current draft of release notes with details on what is completed or incomplete within this project
• [x] Document code purpose, how to use it
  • Mention expected invariants, implicit constraints
• [x] Tests were added for the new behavior
  • Document test purpose, significance of failures
  • Test names should reflect their purpose
• [x] All tests pass (CI will check this if you didn't)
• [x] Serialized types are in stable-versioned modules
• [x] Does this close issues? None