Employers A and B are members in a controlled group. Both have separate 401k plans and separate recordkeepers. Employer A's plan passes coverage using average benefits testing with ratio % barely above safe harbor. Employer B's plan passes coverage easily w/ ratio above 70%.

It has just dawned on one of these employers that they could pass ADP and ACP testing if they permissively aggregate the two plans.

As the employer A recordkeeper, I asked them to send us a copy of the other plan document and noticed some plan differences such as different vesting, different deferral election limit, one has an early retirement age, one has an inservice provision and only one allows for auto rollovers.

Since we're exploring permissive aggregation, it would appear the plan would become subject to BRF testing.
The way I understand BRF, it appears the tests on these differences pass simply because both independent coverage tests have a ratio % greater than the safe harbor. That is, we don't have to do anything special for benefits, rights and features.

Is this correct?

You are assuming that each BRF in each plan is available to all participants in that plan.

I don't know that permissive aggregation goes so far as to render the targeted plan as a part of the tested plan for all reasons. The BRFs you mention are what happens to benefits after they've accrued, yet your permissively aggregating for the purpose of demonstrating nondiscrimination and minimum coverage in the accrual of benefits.