Currently crud spec tests verify that actual and expected results equal by looking at what fields are defined in the actual result and comparing those to the expected fields. This means adding additional fields in a spec test would result in those fields being ignored during verification, and in fact there are several fields which the spec runner does not populate which are thusly ignored.

With the added verifications, one of the bulk writes test started to fail as the driver combined result which failed with a write concern error with the other results, which the spec is not expecting.